Oh - but implementing social programs would cost money - but enforcing quotas is "free"
If you can't keep yourself from getting pregnant, keep the cock out of your vagina.
If you can't support your child, don't get pregnant. See line above.
How hard is it? Why should the rest of the world support your children through social programs so you can "do science"? Be a responsible parent first, and a scientist second.
(For the record, I'm a 26 year old engineer, months away from my PhD, with two children. My wife had the first when I was 22, in college, and we refused what "social program" money was available on principle. I worked two jobs, paid for health insurance, paid for the birth, sold my EverQuest accounts, took 14 semester hours and graduated the same semester, got a real job and got on with life. Sometimes you have to work hard to do what's right, and sometimes that means taking time away from your career and your personal interests to take care of and do whats right by your kids. Deal with it.)
Sure, it's not 1:1 with the final vehicle, do you really expect the first test flight to be?
You've listed the ways it is not Ares I, now list the ways it is not Shuttle. You will find the list is much longer. Yes, only four segments on FS plus a dummy stage but it's not a stock SRB. The upper stage is a mass accurate dummy, but is instrumented for re-entry. The CLV is testing an abort scenario.
It's not a complete PR scenario... some of us are getting * data from this launch.
It's like driving out a Ferrari, but the body is plastic, and there's a Ford engine and a one gear forward only transmission under the hood.
It's more like driving a concept car. It's not street legal, you won't be able to buy it, but it drives and you can get data from it to design the final product you market to the general populace. IMHO.
Where is Ares? Oh, it's in AUTOCAD! Well, that makes ALL the difference!
Guess I'd beg to differ, having seen metal cut for Ares I-X. Just do a google image search and see for yourself.
And by the way, the Ares side of things is, to the best of my knowledge, on schedule to launch in 2009. If you have facts to differ, please let me know. The one thing that will probably delay them is the upcoming Hubble mission - until they vacate pad 39B, the appropriate pad modifications can't be made, so it's a day-for-day slip as the Hubble mission slips.
The LD50 of CO2 for humans is 100,000 PPM (10%), so you'd probably be OK if you were a healthy human with ample oxygen content in the remaining 93%... wouldn't be fun, but tolerable.
The strong force would be the sticky side of the duct tape, and the weak force would be the opposite side of the duct tape, which is still useful but not as strong.
Buy the bun-length hot dogs. Come in 8 packs instead of 10 packs.
But why are you eating hot dogs, when real men eat bratwurst? Now those typically come 5 or 6 on a foam tray, and I have yet to see the bun counterpart.
What is she looking at that is so belabored with content?
The only thing my grandmothers uses their computers for is email. Text. Downloads in seconds. Both tell me that if we grandkids didn't email them as often as we do, they'd throw the damn machines away.
They aren't luddites, just from a different generation. They don't need Twitter streams (fuck, I don't either) or blogs to tell them what's going on in the world. They just want to hear from the grandkids and see what the weather is gonna be outside.
Which of your daily compute tasks is bogging down and could use a boost from multiple CPUs?
Messenger? Word? Email?
I'm an engineer who caps out his quad core Opteron on a daily basis, and submits jobs to queue on a cluster comprised of hundreds of nodes with 2, 4 and 8 processors per node. No, at the moment I can't have enough processing power.
People laugh at primitive statements from pioneers in the industry regarding what is now pitiful amounts of memory being sufficient for future generations.
People will do the same in the future about comments being made today about 1, 2, 4 cores being sufficient for future generations.
It's google. They take your personal information and use it to sell shit to you to make a profit. Of course it doesn't make sense from a privacy stanpoint, but it sure does from a $$$ standpoint.
I'm guessing it doesn't have the "Wow!" factor you were looking to get.
It wasn't for wow factor as you presuppose.
Also, a hell of a lot of the world runs synchronously and if it didn't some very, very bad things would happen.
Sure, but there are a hell of a lot of things in this world running synchronously, in parallel. And many of these separate synchronous events can be influenced by external forces. Nothing exists in a vacuum.
My computer is a tool, like my desk. I'm rarely working on one thing at once, multiple projects across multiple disciplines. Why should my computer be focused on one task? It shouldn't. It should be spread across many tasks. To do this, it can incorporate many parallel processors.
Now take it in context, among other things I write and work with engineering codes for large clusters, modeling and simulation of real-world phenomena. I work with parallel distributed code on a daily basis. This isn't wow factor for me, this is daily life.
Intel provides good support for OpenMP (a standard parallelization technique for shared memory systems, like a desktop) in their compilers, they just think they can do some things better with IntelTBB. I've looked at it, it's neat and it nests with C++ a lot cleaner than OpenMP (which is essentially a bunch of preprocessor commands and not really language integration). However for true scalability (not just on one computer, but distributed computing) you need to parallelize differently using something like MPI.
If you read their pitches on TBB they will say OpenMP might be the better tool for some scenarios, they are pretty honest... this isn't MMX or whatever you are thinking about.
Parallel programming doesn't have to be hard, in fact, it comes very naturally in a number of domains. For example, in finite element analysis (used in a number of math disciplines, including CFD and various stress type calculations) the problem domain is broken down into elements which can naturally be distributed. Calculations within an element are completely independent of the domain until the system of equations are to be solved, and efficient parallelized matrix solvers is old hat.
We got to keep reminding ourselves, the world we live in runs in parallel, why shouldn't our computers?
The only problem would be there are a lot of applications now that have Caps Lock mapped as a default control button... but for people who don't have that as an issue it'd be great:)
It's 100F outside right now, and I wouldn't break a sweat walking to my car.
Someday, I hope... I grew up in Wisconsin, and moved to Alabama. I break a sweat walking to my car in 90F weather. I'm not overweight, I'm in good shape (I bike to work when the weather allows). My body over the last 9 years just hasn't adjusted yet. I keep hoping maybe next summer...
Now winters in Wisconsin, yea, just give me a decent jacket and I'll stay out all day.
My brain is pretty fried right now to do it (long day at work), but you really need two data points:
1. escape velocity of mars
2. distribution of the velocity of the molecules comprising the proposed atmosphere
There are some relatively simple kinetic models for #2 that do a decent enough job. Long story short, if the bulk of the distribution of #2 is greater than #1, then the gas will escape, as it has more velocity than escape velocity. At what rate? Again, depends **how** far above escape the bulk of the distribution is.
Here on earth, the vast bulk of the distribution(s) of each of the consitutents of air fall under the escape velocity of earth - so we lose very little in the way of our atmosphere to space. But we do lose a little here and there. The lower escape velocity on Mars is what hurts its atmosphere potential.
On the other hand, that's not the only natural way to handle the situation you are talking about. Another way is to create a new language. Altogether, I'd rather do the kind of work you're talking about in Matlab or Octave.
I write code for clusters, now specifically CFD. Lots of math, lots of parallel processing. Matlab and Octave isn't gonna cut it, and you really desire the close to the metal aspects of c++.
It may be a narrow range, but there's a lot of people in this narrow range. Specifically, (mechanical/aerospace/etc.) engineers. C++ isn't sexy like Java or Python, but we do a lot of things you just can't do in Python, and can't do fast in Java.
What about using Maple to write the math, and then generating C code from that? better than hacking the language.
Oh, good lord, have you ever seen Maple (or Matlab) C output? it's horrendous. We used to integrate that stuff at work for a project I was on. Never again.
Still need that idiotic space in nested template declarations?
Depends on the compiler. Does not even register as a warning in VC++ (2008). I think it's a warning in GCC 4.2.0. with -Wall. For style reasons I try and leave a space, but it's not necessary.
And it's those myriad idiotic implementation details for why Java and Python
Right, because I really want my code blocks, loops, etc to be defined by space indentation (Python). No choice in style, style is enforced by the programming language. Wait, isn't that the same issue as you were suggesting with C++?
Why should a programming language force me to think about low level implementation details that are nothing to do with the algorithm I'm trying to write?
When you are writing either a video game or cluster software where you are interested in milking out those last few percentile optimizations, and are willing to spend the time and effort thinking about those low-level implementation details to do so. Beyond some of the syntactical advantages C++ has over, say, Java.
Likewise, I think operator overloading is another example of trying to do too much. Yes, it makes programmer classes "first class citizens", but it really has no demonstrable practical benefit in my opinion.
I write math codes for fun and for a living. I have this discussion on an infrequent basis with a Java buddy of mine. Now granted I'm a dumb mechanical engineer and he's a smart CS major, but when I need some custom math classes that aren't provided by the language (tensors, vectors, Jacobians, Quaternions, etc.) and evaluating long math expressions, it is so much easier to view it using the native math symbols than to nest it all in member functions.add(),.subtract(),.multiply(), etc. Once I transliterated a piece of C++ code I had wrote for him and it consumed three times the space and was much less readable due to the nesting of the member functions... being able to use natural math symbols and natural parenthesis makes the math so much more readable, and when 90% of your code is math, you come to appreciate it. I wouldn't have it any other way, at this point.
Slashdot Mirror
Oh - but implementing social programs would cost money - but enforcing quotas is "free"
If you can't keep yourself from getting pregnant, keep the cock out of your vagina.
If you can't support your child, don't get pregnant. See line above.
How hard is it? Why should the rest of the world support your children through social programs so you can "do science"? Be a responsible parent first, and a scientist second.
(For the record, I'm a 26 year old engineer, months away from my PhD, with two children. My wife had the first when I was 22, in college, and we refused what "social program" money was available on principle. I worked two jobs, paid for health insurance, paid for the birth, sold my EverQuest accounts, took 14 semester hours and graduated the same semester, got a real job and got on with life. Sometimes you have to work hard to do what's right, and sometimes that means taking time away from your career and your personal interests to take care of and do whats right by your kids. Deal with it.)
Sure, it's not 1:1 with the final vehicle, do you really expect the first test flight to be?
... some of us are getting * data from this launch.
You've listed the ways it is not Ares I, now list the ways it is not Shuttle. You will find the list is much longer. Yes, only four segments on FS plus a dummy stage but it's not a stock SRB. The upper stage is a mass accurate dummy, but is instrumented for re-entry. The CLV is testing an abort scenario.
It's not a complete PR scenario
It's like driving out a Ferrari, but the body is plastic, and there's a Ford engine and a one gear forward only transmission under the hood.
It's more like driving a concept car. It's not street legal, you won't be able to buy it, but it drives and you can get data from it to design the final product you market to the general populace. IMHO.
Where is Ares? Oh, it's in AUTOCAD! Well, that makes ALL the difference!
Guess I'd beg to differ, having seen metal cut for Ares I-X. Just do a google image search and see for yourself.
And by the way, the Ares side of things is, to the best of my knowledge, on schedule to launch in 2009. If you have facts to differ, please let me know. The one thing that will probably delay them is the upcoming Hubble mission - until they vacate pad 39B, the appropriate pad modifications can't be made, so it's a day-for-day slip as the Hubble mission slips.
The LD50 of CO2 for humans is 100,000 PPM (10%), so you'd probably be OK if you were a healthy human with ample oxygen content in the remaining 93% ... wouldn't be fun, but tolerable.
The strong force would be the sticky side of the duct tape, and the weak force would be the opposite side of the duct tape, which is still useful but not as strong.
I'm not sure if it's possible to kill a cloud with a bullet ...
Buy the bun-length hot dogs. Come in 8 packs instead of 10 packs.
But why are you eating hot dogs, when real men eat bratwurst? Now those typically come 5 or 6 on a foam tray, and I have yet to see the bun counterpart.
What is she looking at that is so belabored with content?
The only thing my grandmothers uses their computers for is email. Text. Downloads in seconds. Both tell me that if we grandkids didn't email them as often as we do, they'd throw the damn machines away.
They aren't luddites, just from a different generation. They don't need Twitter streams (fuck, I don't either) or blogs to tell them what's going on in the world. They just want to hear from the grandkids and see what the weather is gonna be outside.
Which of your daily compute tasks is bogging down and could use a boost from multiple CPUs? Messenger? Word? Email?
I'm an engineer who caps out his quad core Opteron on a daily basis, and submits jobs to queue on a cluster comprised of hundreds of nodes with 2, 4 and 8 processors per node. No, at the moment I can't have enough processing power.
People laugh at primitive statements from pioneers in the industry regarding what is now pitiful amounts of memory being sufficient for future generations.
People will do the same in the future about comments being made today about 1, 2, 4 cores being sufficient for future generations.
wooden instruments sound better the more you play them
... jessica_alba
THAT'S WHAT SHE SAID
It's google. They take your personal information and use it to sell shit to you to make a profit. Of course it doesn't make sense from a privacy stanpoint, but it sure does from a $$$ standpoint.
I'm guessing it doesn't have the "Wow!" factor you were looking to get.
It wasn't for wow factor as you presuppose.
Also, a hell of a lot of the world runs synchronously and if it didn't some very, very bad things would happen.
Sure, but there are a hell of a lot of things in this world running synchronously, in parallel. And many of these separate synchronous events can be influenced by external forces. Nothing exists in a vacuum.
My computer is a tool, like my desk. I'm rarely working on one thing at once, multiple projects across multiple disciplines. Why should my computer be focused on one task? It shouldn't. It should be spread across many tasks. To do this, it can incorporate many parallel processors.
Now take it in context, among other things I write and work with engineering codes for large clusters, modeling and simulation of real-world phenomena. I work with parallel distributed code on a daily basis. This isn't wow factor for me, this is daily life.
Intel provides good support for OpenMP (a standard parallelization technique for shared memory systems, like a desktop) in their compilers, they just think they can do some things better with IntelTBB. I've looked at it, it's neat and it nests with C++ a lot cleaner than OpenMP (which is essentially a bunch of preprocessor commands and not really language integration). However for true scalability (not just on one computer, but distributed computing) you need to parallelize differently using something like MPI.
If you read their pitches on TBB they will say OpenMP might be the better tool for some scenarios, they are pretty honest... this isn't MMX or whatever you are thinking about.
Parallel programming doesn't have to be hard, in fact, it comes very naturally in a number of domains. For example, in finite element analysis (used in a number of math disciplines, including CFD and various stress type calculations) the problem domain is broken down into elements which can naturally be distributed. Calculations within an element are completely independent of the domain until the system of equations are to be solved, and efficient parallelized matrix solvers is old hat.
We got to keep reminding ourselves, the world we live in runs in parallel, why shouldn't our computers?
Our world runs in parallel; why shouldn't our computers?
64 sounds good to me. I wouldn't complain if I had more.
The only problem would be there are a lot of applications now that have Caps Lock mapped as a default control button ... but for people who don't have that as an issue it'd be great :)
Tony Robbins Hungry !!!
Yeah, I like to go baby seal hunting with a baseball bat.
But it's cool. I buy baby seal offsets.
It's 100F outside right now, and I wouldn't break a sweat walking to my car.
... I grew up in Wisconsin, and moved to Alabama. I break a sweat walking to my car in 90F weather. I'm not overweight, I'm in good shape (I bike to work when the weather allows). My body over the last 9 years just hasn't adjusted yet. I keep hoping maybe next summer ...
Someday, I hope
Now winters in Wisconsin, yea, just give me a decent jacket and I'll stay out all day.
My brain is pretty fried right now to do it (long day at work), but you really need two data points:
1. escape velocity of mars
2. distribution of the velocity of the molecules comprising the proposed atmosphere
There are some relatively simple kinetic models for #2 that do a decent enough job. Long story short, if the bulk of the distribution of #2 is greater than #1, then the gas will escape, as it has more velocity than escape velocity. At what rate? Again, depends **how** far above escape the bulk of the distribution is.
Here on earth, the vast bulk of the distribution(s) of each of the consitutents of air fall under the escape velocity of earth - so we lose very little in the way of our atmosphere to space. But we do lose a little here and there. The lower escape velocity on Mars is what hurts its atmosphere potential.
On the other hand, that's not the only natural way to handle the situation you are talking about. Another way is to create a new language. Altogether, I'd rather do the kind of work you're talking about in Matlab or Octave.
I write code for clusters, now specifically CFD. Lots of math, lots of parallel processing. Matlab and Octave isn't gonna cut it, and you really desire the close to the metal aspects of c++.
It may be a narrow range, but there's a lot of people in this narrow range. Specifically, (mechanical/aerospace/etc.) engineers. C++ isn't sexy like Java or Python, but we do a lot of things you just can't do in Python, and can't do fast in Java.
What about using Maple to write the math, and then generating C code from that? better than hacking the language.
Oh, good lord, have you ever seen Maple (or Matlab) C output? it's horrendous. We used to integrate that stuff at work for a project I was on. Never again.
Still need that idiotic space in nested template declarations?
Depends on the compiler. Does not even register as a warning in VC++ (2008). I think it's a warning in GCC 4.2.0. with -Wall. For style reasons I try and leave a space, but it's not necessary.
And it's those myriad idiotic implementation details for why Java and Python
Right, because I really want my code blocks, loops, etc to be defined by space indentation (Python). No choice in style, style is enforced by the programming language. Wait, isn't that the same issue as you were suggesting with C++?
Why should a programming language force me to think about low level implementation details that are nothing to do with the algorithm I'm trying to write?
When you are writing either a video game or cluster software where you are interested in milking out those last few percentile optimizations, and are willing to spend the time and effort thinking about those low-level implementation details to do so. Beyond some of the syntactical advantages C++ has over, say, Java.
Likewise, I think operator overloading is another example of trying to do too much. Yes, it makes programmer classes "first class citizens", but it really has no demonstrable practical benefit in my opinion.
.add(), .subtract(), .multiply(), etc. Once I transliterated a piece of C++ code I had wrote for him and it consumed three times the space and was much less readable due to the nesting of the member functions... being able to use natural math symbols and natural parenthesis makes the math so much more readable, and when 90% of your code is math, you come to appreciate it. I wouldn't have it any other way, at this point.
I write math codes for fun and for a living. I have this discussion on an infrequent basis with a Java buddy of mine. Now granted I'm a dumb mechanical engineer and he's a smart CS major, but when I need some custom math classes that aren't provided by the language (tensors, vectors, Jacobians, Quaternions, etc.) and evaluating long math expressions, it is so much easier to view it using the native math symbols than to nest it all in member functions