Catholic] Church's historical position on the immobility of the Earth was not only scientifically supportable, but it was the most stable model of the universe
The Roman Catholic Church long ago accepted our current scientific understanding of the organization of celestial bodies.
Oh, and evolution through natural selection as well.
And one of its greatest thinkers believed that reason and faith were both equally valid ways to truth and not in conflict at all.
These nuts are in no way affiliated with official Roman Catholic Church positions. So let's just halt the Church bashing before we begin, ok?
In the first example, 'm' is not being moved to the constant data section. The constant vector being assigned to m is placed there. MSVC is missing the vectorization, not placement of constants into constant memory. You can see that it fetches the constant values from memory using scalar moves while gcc and icc use vector moves.
I'm not familiar with MSVC switches but you might need to tell it explicitly to vectorize. I'm curious why you didn't try -ftree-vectorize with gcc, for example.
Floating-point optimization is a tricky thing. Many compilers will be very conservative to retain bitwise equivalent results regardless of optimization level. Some will even go as far as maintaining bitwise equivalence between scalar and vector code. That can severely degrade optimization. Again, most compilers have a switch to enable "unsafe" floating-point optimization. This may be what's tripping up these compilers in some cases.
NaNs are also an issue with floating-point. The compiler is not allowed to eliminate anything which might raise an exception.
When encountering intrinsics, many compilers will do exactly as you say, as noted in the article. That's not a bug, it's a feature. When people use intrinsics, they usually are trying to hand-code something and often don't want the compiler to mess with it.
Some of these tests (the shuffle one for example) are a little out-of-the-ordinary. Compiler developer time is at a premium and it's not worth doing these kinds of micro-optimizations if such code is never seen in the wild. That said, it's clear the some compilers (gcc, for example, and LLVM) do these sorts of things.
On x86, it's often just fine to spill things to the stack and reload them. My studies show that the number of spills does not matter so much but rather what is spilled. So the number of loads/stores, while a gross indicator of performance, doesn't tell the whole story.
The comparison test is, I think, one of those cases not worth optimizing. I can't recall ever seeing a vector compare where the operands are known statically. Doing that optimization would require loading static vectors of various combinations of 1s and 0s from memory. It is almost certainly faster to just do the compare. This isn't a missed optimization. In gcc's case it's the compiler doing what it should, regardless of what the programmer expects.
Even so, these are interesting code examples. It would be neat to see what happens when we turn on -ftree-vectorize, use a newer gcc or try LLVM.
This would be a good discussion to have. I have seen many programmers attempt to "help" the compiler only to make some previous trivial transformation impossible. Can you post some code examples? I think it would be a great educational opportunity.
In some fields, the solutions will have to be in software. With some massively parallel codes, even a 2x speedup in interconnect speed won't help that much. It's analogous to trying to speed up an exponential algorithms by increasing clock speed. In the end, it just doesn't matter. Faster interconnect will help a lot of things but we're going to have to bite the bullet and learn how to program again.
CPU interconnect bandwidth and memory bandwidth is now large enough that this is no longer an issue
Well, it depends on what kinds of codes you're running. Memory bandwidth is becoming a bigger problem, not a smaller one. Communication overhead can easily dominate in large parallel codes. These are not niche things, either. We're going more parallel, not less.
I think you missed the point entirely. It is "really that hard" for some people to move up in society, because there are all sorts of things in the way. Tell the black contractor how easy it is just after he was passed over for an abatement job he was certified to do in favor of some white contractor who needed funding for the extra training to do the job (true story, I know the guy). The system is rigged. It may not be rigged "intentionally" in the sense of being a grand planned strategy but it is rigged nonetheless.
One may be in the country illegally, but a person cannot be "illegal."
And waiting in line for hundreds of years is a little impractical, wouldn't you say? And yes, according to official immigration policy, that's how long it would take for some people to "follow the rules." In that case, civil disobedience is entirely appropriate.
Your answer is far too simplistic. There are many systemic barriers in the way of climbing social status in the U.S. If one is well-off, it's likely the children will be too, not because of how hard they or their children worked but because the system is set up to make that the default outcome. That's not to trivialize the work they or their children do but the poor and non-whites need to work much, much harder to achieve the same result as a white or well-off person. That's just simple fact, demonstrated repeatedly by various societal studies.
And the most hours. That's because they have to overcome a lot of systemic barriers. If you are poor or non-white, you have to work much, much harder to climb social status in the U.S.
The dilemma I'm having is the need to run servers. I host web and mail and the standard companies all have user agreements prohibiting that. So I'm stuck with an expensive DSL when I'd really like to upgrade to something faster.
Anyone have suggestions?
I am not claiming that a tax rate necessarily causes posperity. I'm arguing that it doesn't seem to hurt it. I'll also argue that investment in infrastructure pays huge dividends and we need taxes to do that. More taxes than we have today. Our infrastructure is falling apart because we haven't kept up maintenance, let alone improved it.
The mortgage interest deduction should be abolished. It is another example of systemic racism in that the primary beneficiaries are whites. We need a much more progressive tax system, including the income tax.
If your household brings in ~$158k you're already in the top 5%.
If anything we need to lower the threshold at which a higher income tax kicks in. Income is severely undertaxed in the U.S. Back in the 1950's (one of the most prosperous times in our history) the highest income tax rate was at 92%. This is one of the direct causes of tyhe severe wealth disparity we have today, and it's getting worse. We can't survive as a society if this continues.
I think its more a case of some people perceiving that their own interests coincide with those of the moneyed classes. Its not that they're being hoodwinked so much as that they're making a similar analysis and attempting to strengthen their own position with half-true arguments.
It seems more a case of people identifying with the rich because they believe they can be rich someday. The fact that our current systemic barriers and biases will prevent that for the vast majority of people is not something people want to think about. The myth of rugged individualism is a strong one.
29% of U.S. greenhouse gas emissions comes from transportation and it is the fastest-growing source of GHGs as explained here. So yes, the SUV does matter. A lot.
I struggle to find a decent use of them and often see them used where the implementation limits future maintenance. I know in C++ templateing was a huge? feature, which generics seem similar to. I don't do any of C++ work so maybe I'm just missing the point but I struggle to see much value in it.
I don't do any Java work but code C++ every day. Generics are hugely useful if done right. All I ever hear about Java is that generics are for containers. That's actually a very restrictive model. In C++ generics applies to everything. It gives you duck typing with static type checking. It gives you a compile-time functional programming model with pattern matching.
It gives you static polymorphism with very loose ("textual" is how I would describe it) binding, which means I can write a function that operates on any combination of types and it will work as long as those types implement the textual interfaces that function requires (+, *, print(), and so on). Those types need not have any relationship to each other at all, which means inheritance is much less important than it used to be. This in turn makes code even more reusable.
Aside from the error message issues with C++ templates (and that's a big issue, no question), they are an excellent way to write code that is both reusable and high-performance.
For example, I love duck typing. I can write a class that implements read() and pass it to almost any function that expects to receive a file-type object as an argument. I can write def add(x,y): return x+y and know that it will do the right thing whether I pass in ints, floats, or strings.
This is indeed convenient, but the functionality is not limited to dynamically-typed languages. C++ templates provide the same benefits with the bonus of static type checking. I'm curious to hear your opinion on static typing and generic programming and whether something like C++ templates provides what you're looking for. Is there something else about dynamic typing that you find useful?
A dynamic language like SmallTalk can be nice because of introspection and the ability to modify types at runtime. A statically-typed language can't do that. I wouldn't want to use such introspection and type modification in big projects, though.
So, come on, explain to my why this is (allegedly) so.
"Fewer" is to refer to a quantity that can be counted or is discrete ("fewer stones") while "less" is to refer to a quantity that can't easily be counted or is continuous ("less water"). Compare "less sugar" to "fewer sugar granules."
And the real solution is to have one line with people directed to registers as they open up. Like practically every other queuing system out there (banks, DMV, etc.). Why retail gets this so wrong is beyond me.
Exactly - there's always a presumption of guilt for the man.
False. I know people who have been raped and had their employer/school/etc. do whatever it took to cover it up. I've know prosecutors to refuse to prosecute because of the position/importance of said institution in the community.
Class is often used as a substitute to avoid talking about race. Economic class is another divider, for sure, but race cannot be ignored. Racial segregation and discrimination is completely embedded into U.S. society and we have to actively counter it with policies that explicitly try to undo it. That's what the Kirwin people are getting at with the quote you cited. It's not enough for policy to be colorblind. That would simply maintain the status quo.
The truth is that the 20% and the 1% are almost completely white. That's not a coincidence. And that same group of privileged people use race as a divider to keep the unwealthy from working together. Witness the birther movement, the tea party and the racial undertones are Fox News talking heads. This is not accidental. None of these things is completely grass-roots. They are very much organized and funded by the rich and powerful.
I agree with a lot of what you're saying. Choices are important, but the way our society is structured often limits them. For example:
Dr. Fuhrman, for example, has built an eating plan that works to reduce lots of disease, based on thousands of scientific studies that say nutrition is a very significant aspect of health:
Yes, nutrition is very important. That's why where you live really matters. Do you have easy access to fresh foods and vegetables? Is there are farmer's market near you? Or is it a sea of fast-food restaurants? It turns out that in poor and segregated neighborhoods, the most convenient food sources are processed sugary foods from quick-stop gas stations and fast food joints. It's pretty tough to find a farmer's market or organic co-op in poor or racially segregated communities.
Slashdot Mirror
The summary should read:
Catholic] Church's historical position on the immobility of the Earth was not only scientifically supportable, but it was the most stable model of the universe
The Roman Catholic Church long ago accepted our current scientific understanding of the organization of celestial bodies.
Oh, and evolution through natural selection as well.
And one of its greatest thinkers believed that reason and faith were both equally valid ways to truth and not in conflict at all.
These nuts are in no way affiliated with official Roman Catholic Church positions. So let's just halt the Church bashing before we begin, ok?
A couple of things:
In the first example, 'm' is not being moved to the constant data section. The constant vector being assigned to m is placed there. MSVC is missing the vectorization, not placement of constants into constant memory. You can see that it fetches the constant values from memory using scalar moves while gcc and icc use vector moves.
I'm not familiar with MSVC switches but you might need to tell it explicitly to vectorize. I'm curious why you didn't try -ftree-vectorize with gcc, for example.
Floating-point optimization is a tricky thing. Many compilers will be very conservative to retain bitwise equivalent results regardless of optimization level. Some will even go as far as maintaining bitwise equivalence between scalar and vector code. That can severely degrade optimization. Again, most compilers have a switch to enable "unsafe" floating-point optimization. This may be what's tripping up these compilers in some cases.
NaNs are also an issue with floating-point. The compiler is not allowed to eliminate anything which might raise an exception.
When encountering intrinsics, many compilers will do exactly as you say, as noted in the article. That's not a bug, it's a feature. When people use intrinsics, they usually are trying to hand-code something and often don't want the compiler to mess with it.
Some of these tests (the shuffle one for example) are a little out-of-the-ordinary. Compiler developer time is at a premium and it's not worth doing these kinds of micro-optimizations if such code is never seen in the wild. That said, it's clear the some compilers (gcc, for example, and LLVM) do these sorts of things.
On x86, it's often just fine to spill things to the stack and reload them. My studies show that the number of spills does not matter so much but rather what is spilled. So the number of loads/stores, while a gross indicator of performance, doesn't tell the whole story.
The comparison test is, I think, one of those cases not worth optimizing. I can't recall ever seeing a vector compare where the operands are known statically. Doing that optimization would require loading static vectors of various combinations of 1s and 0s from memory. It is almost certainly faster to just do the compare. This isn't a missed optimization. In gcc's case it's the compiler doing what it should, regardless of what the programmer expects.
Even so, these are interesting code examples. It would be neat to see what happens when we turn on -ftree-vectorize, use a newer gcc or try LLVM.
Production-level compilers haven't been written in assembler for at least a couple of decades.
This would be a good discussion to have. I have seen many programmers attempt to "help" the compiler only to make some previous trivial transformation impossible. Can you post some code examples? I think it would be a great educational opportunity.
In some fields, the solutions will have to be in software. With some massively parallel codes, even a 2x speedup in interconnect speed won't help that much. It's analogous to trying to speed up an exponential algorithms by increasing clock speed. In the end, it just doesn't matter. Faster interconnect will help a lot of things but we're going to have to bite the bullet and learn how to program again.
CPU interconnect bandwidth and memory bandwidth is now large enough that this is no longer an issue
Well, it depends on what kinds of codes you're running. Memory bandwidth is becoming a bigger problem, not a smaller one. Communication overhead can easily dominate in large parallel codes. These are not niche things, either. We're going more parallel, not less.
I think you missed the point entirely. It is "really that hard" for some people to move up in society, because there are all sorts of things in the way. Tell the black contractor how easy it is just after he was passed over for an abatement job he was certified to do in favor of some white contractor who needed funding for the extra training to do the job (true story, I know the guy). The system is rigged. It may not be rigged "intentionally" in the sense of being a grand planned strategy but it is rigged nonetheless.
What?
One may be in the country illegally, but a person cannot be "illegal."
And waiting in line for hundreds of years is a little impractical, wouldn't you say? And yes, according to official immigration policy, that's how long it would take for some people to "follow the rules." In that case, civil disobedience is entirely appropriate.
Your answer is far too simplistic. There are many systemic barriers in the way of climbing social status in the U.S. If one is well-off, it's likely the children will be too, not because of how hard they or their children worked but because the system is set up to make that the default outcome. That's not to trivialize the work they or their children do but the poor and non-whites need to work much, much harder to achieve the same result as a white or well-off person. That's just simple fact, demonstrated repeatedly by various societal studies.
And the most hours. That's because they have to overcome a lot of systemic barriers. If you are poor or non-white, you have to work much, much harder to climb social status in the U.S.
Except how can a person be illegal? That's what I don't understand.
The dilemma I'm having is the need to run servers. I host web and mail and the standard companies all have user agreements prohibiting that. So I'm stuck with an expensive DSL when I'd really like to upgrade to something faster. Anyone have suggestions?
I am not claiming that a tax rate necessarily causes posperity. I'm arguing that it doesn't seem to hurt it. I'll also argue that investment in infrastructure pays huge dividends and we need taxes to do that. More taxes than we have today. Our infrastructure is falling apart because we haven't kept up maintenance, let alone improved it.
The mortgage interest deduction should be abolished. It is another example of systemic racism in that the primary beneficiaries are whites. We need a much more progressive tax system, including the income tax.
People who earn $250,000 a year are not rich
Yes, they are. People who think they aren't are really disconnected from reality.
http://en.wikipedia.org/wiki/Household_income_in_the_United_States
If your household brings in ~$158k you're already in the top 5%.
If anything we need to lower the threshold at which a higher income tax kicks in. Income is severely undertaxed in the U.S. Back in the 1950's (one of the most prosperous times in our history) the highest income tax rate was at 92%. This is one of the direct causes of tyhe severe wealth disparity we have today, and it's getting worse. We can't survive as a society if this continues.
I think its more a case of some people perceiving that their own interests coincide with those of the moneyed classes. Its not that they're being hoodwinked so much as that they're making a similar analysis and attempting to strengthen their own position with half-true arguments.
It seems more a case of people identifying with the rich because they believe they can be rich someday. The fact that our current systemic barriers and biases will prevent that for the vast majority of people is not something people want to think about. The myth of rugged individualism is a strong one.
29% of U.S. greenhouse gas emissions comes from transportation and it is the fastest-growing source of GHGs as explained here. So yes, the SUV does matter. A lot.
I struggle to find a decent use of them and often see them used where the implementation limits future maintenance. I know in C++ templateing was a huge? feature, which generics seem similar to. I don't do any of C++ work so maybe I'm just missing the point but I struggle to see much value in it.
I don't do any Java work but code C++ every day. Generics are hugely useful if done right. All I ever hear about Java is that generics are for containers. That's actually a very restrictive model. In C++ generics applies to everything. It gives you duck typing with static type checking. It gives you a compile-time functional programming model with pattern matching.
It gives you static polymorphism with very loose ("textual" is how I would describe it) binding, which means I can write a function that operates on any combination of types and it will work as long as those types implement the textual interfaces that function requires (+, *, print(), and so on). Those types need not have any relationship to each other at all, which means inheritance is much less important than it used to be. This in turn makes code even more reusable.
Aside from the error message issues with C++ templates (and that's a big issue, no question), they are an excellent way to write code that is both reusable and high-performance.
For example, I love duck typing. I can write a class that implements read() and pass it to almost any function that expects to receive a file-type object as an argument. I can write def add(x,y): return x+y and know that it will do the right thing whether I pass in ints, floats, or strings.
This is indeed convenient, but the functionality is not limited to dynamically-typed languages. C++ templates provide the same benefits with the bonus of static type checking. I'm curious to hear your opinion on static typing and generic programming and whether something like C++ templates provides what you're looking for. Is there something else about dynamic typing that you find useful?
A dynamic language like SmallTalk can be nice because of introspection and the ability to modify types at runtime. A statically-typed language can't do that. I wouldn't want to use such introspection and type modification in big projects, though.
So, come on, explain to my why this is (allegedly) so.
"Fewer" is to refer to a quantity that can be counted or is discrete ("fewer stones") while "less" is to refer to a quantity that can't easily be counted or is continuous ("less water"). Compare "less sugar" to "fewer sugar granules."
And the real solution is to have one line with people directed to registers as they open up. Like practically every other queuing system out there (banks, DMV, etc.). Why retail gets this so wrong is beyond me.
Thank you. I'm glad some people are willing to stand up for what's right.
Exactly - there's always a presumption of guilt for the man.
False. I know people who have been raped and had their employer/school/etc. do whatever it took to cover it up. I've know prosecutors to refuse to prosecute because of the position/importance of said institution in the community.
Class is often used as a substitute to avoid talking about race. Economic class is another divider, for sure, but race cannot be ignored. Racial segregation and discrimination is completely embedded into U.S. society and we have to actively counter it with policies that explicitly try to undo it. That's what the Kirwin people are getting at with the quote you cited. It's not enough for policy to be colorblind. That would simply maintain the status quo.
The truth is that the 20% and the 1% are almost completely white. That's not a coincidence. And that same group of privileged people use race as a divider to keep the unwealthy from working together. Witness the birther movement, the tea party and the racial undertones are Fox News talking heads. This is not accidental. None of these things is completely grass-roots. They are very much organized and funded by the rich and powerful.
I agree with a lot of what you're saying. Choices are important, but the way our society is structured often limits them. For example:
Dr. Fuhrman, for example, has built an eating plan that works to reduce lots of disease, based on thousands of scientific studies that say nutrition is a very significant aspect of health:
Yes, nutrition is very important. That's why where you live really matters. Do you have easy access to fresh foods and vegetables? Is there are farmer's market near you? Or is it a sea of fast-food restaurants? It turns out that in poor and segregated neighborhoods, the most convenient food sources are processed sugary foods from quick-stop gas stations and fast food joints. It's pretty tough to find a farmer's market or organic co-op in poor or racially segregated communities.