The FED just issued another $60 billion auction to keep the banks alive. This thing won't end until someone figures out what to do with all these lousy loans.
The scariest thing is that a jury is deciding these cases. After filing a patent and having the lawyer rewrite it into Patentese I could barely understand it myself. How can 12 average joes decide a technical matter like a patent trial?
In my opinion, it's marketing that screws the tech of MS
Don't let the Microsoft Software Department off the hook. They are a total disaster and are completely incompetent. Everyone knows Vista is a total disaster, so I'll focus on their other technologies.
First off, it is a pain the ass to write anything for Windows. You go out and learn MFC. Then they tell you, "Oh no!! Don't do that! We want you to go learn our new language, C#, and learn Windows Forms." Before you are done with that you get, "No! Windows Forms is the old way of doing things. We want you to learn WPF and XAML!"
I mainly want Microsoft to fail so that I never need to deal with any more of their bad technology and horrible standards.
I've got to agree with this. Look at how many people in industry get bitter and angry when someone expects them to know how to sort or know some math. Parallel programming is not difficult. Just too few engineers care about it since they are too busy learning some new fad: the latest scripting language, some new OOP design pattern, etc. No university should allow anyone to graduate with a CS degree without having taken some parallel programming theory and lab classes.
That Constitution says nothing about the WTO getting to change US laws they don't like
Exactly! The Constitution also spells out that the Supreme Court is the highest court in the land. Congress and the president don't have the legal right to give jurisdiction to a foreign court.
And as an aside, what a dumb law that Irish Music thing is. That's nice, you can't play the radio at work without forking money over to somebody.
How? Unless he stole source code and used it in his program, I don't see how.
They are claiming that the tool makes a copy of the game and stores it to ram to avoid their anti-cheating checks. Interesting to see if it is illegal to make a temporary copy (for your own personal use) of a program you legally purchased.
It doesn't sound as bad as I thought from the title of the article. Seems they are just going to use it for tactical situations. So if there is a hostage situation, they can send up one of these things over the area to get a better view of the situation. Seems pretty useful: if you are sending in a SWAT team, you could quickly notify them if someone with a gun jumped out a window and is hiding in some bushes.
The only danger is that they decide to expand the program and start having these things all over the place. Or what if they use them to videotape people peacefully protesting to get a list of "trouble makers" for the FBI to keep tabs on.
I disagree with you about the need for patents. I'll admit that most patents are pretty obvious. But every now and then someone works hard for something that really is unique. Without patents, the result will be predictable: most people will keep their algorithms a closely guarded secret. The result will be that academia will suffer as algorithms go from publicly disclosed patents to trade secrets. A better idea is for the patent offices to be less generous in what they are willing to grant patents for.
I'm often amazed how many people I meet at work that call themselves engineers but don't know how to write an equation for a line given a couple data points. (If you've ever worked with analog to digital converters you'll understand why this is a basic skill.) I guess I shouldn't be surprised when I constantly hear stuff about "Why do I need to know all this math and theory? I want to be a programmer!" I know a lot of guys who were physics or math majors and are better programmers than the CS majors because of that attitude. They don't write pretty-looking object oriented programs, but they write programs that work and are more efficient.
When dealing with problems like this, the value comes from just having a proof. This sounds a little weird, but the tools that are used to prove something are more valuable than the theorems in many cases. I'll list some examples of why this is the case.
Cantor gave a famous proof that the real numbers were uncountable (in laymen's terms you can think of this as proving that the set of real numbers is much larger than the set of integers.) In order to do this, he invented a technique called "diagonalization." This method was used to produce much of modern logic and set theory: Godel used it for his famous Incompleteness Theorem. The Halting Problem relies on diagonalization as well.
When Joseph Cohen solved the Continuum Hypothesis, he developed a technique called Forcing. This technique has made many difficult problems in Set Theory solvable.
So in short, the hope is that a proof will provide new insights and new techniques that will open up many new doors.
This is why so many people look down on philosophy: it runs counter to common sense.
Following this train of logic, when I'm buying a new car I should ignore that the salesman only makes money if he sells me a car. So when he's busy telling me that the 1982 Volkswagon he's trying to sell me could out-accelerate a Porsche, I should just treat it as an impartial opinion
The poster is completely correct in pointing out that an IBM representative has an inherent bias against a Microsoft standard and it's wrong to label his post as a flame.
Basically VHDL is from the ADA/Pascal family of languages. And Verilog looks like old K&R C.
VHDL is used mostly in defense industries and outside the United States. Everywhere else, Verilog is more popular. Nothing you could do in either language that you couldn't do in the other. But the VHDL syntax looks a lot cleaner in my opinion. But the price to pay for this is that VHDL is a fatter language with many more constructs (it even has pointers--to make modeling easier.)
I think the problem is people don't realize what an extremely difficult problem programming is. Perhaps it helps to state it mathematically: Given a mathematical specification of a program (i.e. we might say "Given N is positive, set x to the square root of N") it is not possible to construct an algorithm to construct a program from that specification (for the general case.) Furthermore, it is not even possible to tell, in the general case, whether there is even a solution.
So from a mathematical standpoint, programming is a difficult and creative process. And it's no surprise that it will always be buggy.
Agreed, but because of the money involved, the Pac's congressmen listen to are the ones that represent a small minority most people disagree with. It's no accident that people can't buy their prescription drugs from another country more cheaply - even though everyone in Congress is such a free-trade fan.
Thanks for the answer. I find it interesting that the hard problem seems to be to filter out irrelevant data. I work on digital signal processing hardware and we have a similar problem. Most of the hardware isn't even working on the real problem: most of it is for reducing noise in the system and ignoring short random spikes. Weird that the human brain is still the best garbage filter ever designed.
I'm not even sure it is actually passing the full Turing test. From the article I noticed one of the researchers saying: "That's how we plan to pass this limited version of the Turing test." Anyone know what he means by this being a "limited" version of the Turing test?
I'm not quite sure why you need all of this excessive math and science Depends on what you are doing. If you are working on web apps and some Open Source projects you can get away with it. But start doing some digital signal processing/ encryption and you'll find out that functional analysis, abstract algebra, and the Fourier transform are important tools.
So the best thing is to not give asshats ways to loot billions of dollars from society in general. I haven't read up to see if this is a real patent or an ambush (patent some easy idea- let it become popular by hiding the patent- then ambush). It sounds like an ambush tho. You make a valid point about stupid patents. I'm sure every engineer has seen patents that got them to roar with laughter at how obvious they are. But the fix isn't to have bribed senators use eminent domain and billion$ of tax dollars to swipe a company's property. The fix is to get the patent office to stop being so liberal in the granting of patents. I doubt that what they are doing is even constitutional. The Constitution sets up clear property rights on patents. There is no escape clause anywhere in it that allows the Congress to pass laws nullifying a patent.
Countries like China and Russia will spend money on weapons where they ought to be feeding, clothing, and providing to the medical care of their kids. I understand what you're saying but military spending actually does get people clothed and fed. Look at what military spending has produced: televisions, HD-tv, computers, semiconductors, atomic energy, quantum physics, the internet, jet engines, GPS, etc. As contradictory as it may seem, most of the technology that makes people's lives better was actually originally designed for military purposes - non-military research yields far fewer fruits. I'm all for a smaller military but not for smaller military research budgets.
That's kind of my point. The market here is being filled domestically because it's a crappy career choice. When people were being paid well and treated well in the late 90's people were flocking to engineering. Now there are less and less engineering majors because people know they are much better off with a medical degree, law degree, or going into finance.
This is bullshit. If foreigners are so smart, why do they have to come to the US for jobs? Why are people surprised noone wants to go into engineering in the US: stagnant wages, offshoring, age discrimination, long hours. It's a shitty way to waste $100k on an education.
I was starting to get the signs of carpal tunnel: I was losing feeling in my left thumb. For all of you younger engineers, losing sensation in your fingers is the first sign of early carpal tunnel syndrome. I bought a Kinesis for about $300, and my hands healed within a couple weeks. It's a good 5 years later and I am still happy with my investment.
I do this all the time on decent sized projects. I am mostly working on control loops. For most things I will do informal proofs. For some very critical sections, I will go the full route and do a full proof. And yes, it does work. Spending an extra couple hours proving things is time well-spent when a full circuit compile will take me 2 hours. And yeah, if you use formal methods, except for syntax errors 90% of stuff works perfectly on the first run. One section of code was taking up too much chip area. I used a formal method and cut the circuit size 80% with no loss in speed. So finally, formal methods do produce leaner and more bug-free code. More success stories from a simple google search http://shemesh.larc.nasa.gov/fm/fm-what.html.
I have a strong suspicion that you have never worked on an embedded system of any size.
There are a lot of heinous-looking things that pop up in embedded software that are absolutely necessary to get around hardware problems, timing issues, etc.
Try refactoring that stuff out and see how fast the whole system will crash into a pile of rubble.
Slashdot Mirror
The FED just issued another $60 billion auction to keep the banks alive. This thing won't end until someone figures out what to do with all these lousy loans.
The scariest thing is that a jury is deciding these cases. After filing a patent and having the lawyer rewrite it into Patentese I could barely understand it myself. How can 12 average joes decide a technical matter like a patent trial?
Don't let the Microsoft Software Department off the hook. They are a total disaster and are completely incompetent. Everyone knows Vista is a total disaster, so I'll focus on their other technologies.
First off, it is a pain the ass to write anything for Windows. You go out and learn MFC. Then they tell you, "Oh no!! Don't do that! We want you to go learn our new language, C#, and learn Windows Forms." Before you are done with that you get, "No! Windows Forms is the old way of doing things. We want you to learn WPF and XAML!"
I mainly want Microsoft to fail so that I never need to deal with any more of their bad technology and horrible standards.
I've got to agree with this. Look at how many people in industry get bitter and angry when someone expects them to know how to sort or know some math. Parallel programming is not difficult. Just too few engineers care about it since they are too busy learning some new fad: the latest scripting language, some new OOP design pattern, etc. No university should allow anyone to graduate with a CS degree without having taken some parallel programming theory and lab classes.
Exactly! The Constitution also spells out that the Supreme Court is the highest court in the land. Congress and the president don't have the legal right to give jurisdiction to a foreign court.
And as an aside, what a dumb law that Irish Music thing is. That's nice, you can't play the radio at work without forking money over to somebody.
They are claiming that the tool makes a copy of the game and stores it to ram to avoid their anti-cheating checks. Interesting to see if it is illegal to make a temporary copy (for your own personal use) of a program you legally purchased.
It doesn't sound as bad as I thought from the title of the article. Seems they are just going to use it for tactical situations. So if there is a hostage situation, they can send up one of these things over the area to get a better view of the situation. Seems pretty useful: if you are sending in a SWAT team, you could quickly notify them if someone with a gun jumped out a window and is hiding in some bushes.
The only danger is that they decide to expand the program and start having these things all over the place. Or what if they use them to videotape people peacefully protesting to get a list of "trouble makers" for the FBI to keep tabs on.
I disagree with you about the need for patents. I'll admit that most patents are pretty obvious. But every now and then someone works hard for something that really is unique. Without patents, the result will be predictable: most people will keep their algorithms a closely guarded secret. The result will be that academia will suffer as algorithms go from publicly disclosed patents to trade secrets. A better idea is for the patent offices to be less generous in what they are willing to grant patents for.
I agree with you completely.
I'm often amazed how many people I meet at work that call themselves engineers but don't know how to write an equation for a line given a couple data points. (If you've ever worked with analog to digital converters you'll understand why this is a basic skill.) I guess I shouldn't be surprised when I constantly hear stuff about "Why do I need to know all this math and theory? I want to be a programmer!" I know a lot of guys who were physics or math majors and are better programmers than the CS majors because of that attitude. They don't write pretty-looking object oriented programs, but they write programs that work and are more efficient.
When dealing with problems like this, the value comes from just having a proof. This sounds a little weird, but the tools that are used to prove something are more valuable than the theorems in many cases. I'll list some examples of why this is the case.
Cantor gave a famous proof that the real numbers were uncountable (in laymen's terms you can think of this as proving that the set of real numbers is much larger than the set of integers.) In order to do this, he invented a technique called "diagonalization." This method was used to produce much of modern logic and set theory: Godel used it for his famous Incompleteness Theorem. The Halting Problem relies on diagonalization as well.
When Joseph Cohen solved the Continuum Hypothesis, he developed a technique called Forcing. This technique has made many difficult problems in Set Theory solvable.
So in short, the hope is that a proof will provide new insights and new techniques that will open up many new doors.
This is why so many people look down on philosophy: it runs counter to common sense.
Following this train of logic, when I'm buying a new car I should ignore that the salesman only makes money if he sells me a car. So when he's busy telling me that the 1982 Volkswagon he's trying to sell me could out-accelerate a Porsche, I should just treat it as an impartial opinion
The poster is completely correct in pointing out that an IBM representative has an inherent bias against a Microsoft standard and it's wrong to label his post as a flame.
Basically VHDL is from the ADA/Pascal family of languages. And Verilog looks like old K&R C.
VHDL is used mostly in defense industries and outside the United States. Everywhere else, Verilog is more popular. Nothing you could do in either language that you couldn't do in the other. But the VHDL syntax looks a lot cleaner in my opinion. But the price to pay for this is that VHDL is a fatter language with many more constructs (it even has pointers--to make modeling easier.)
I think the problem is people don't realize what an extremely difficult problem programming is. Perhaps it helps to state it mathematically: Given a mathematical specification of a program (i.e. we might say "Given N is positive, set x to the square root of N") it is not possible to construct an algorithm to construct a program from that specification (for the general case.) Furthermore, it is not even possible to tell, in the general case, whether there is even a solution.
So from a mathematical standpoint, programming is a difficult and creative process. And it's no surprise that it will always be buggy.
Agreed, but because of the money involved, the Pac's congressmen listen to are the ones that represent a small minority most people disagree with. It's no accident that people can't buy their prescription drugs from another country more cheaply - even though everyone in Congress is such a free-trade fan.
Thanks for the answer. I find it interesting that the hard problem seems to be to filter out irrelevant data. I work on digital signal processing hardware and we have a similar problem. Most of the hardware isn't even working on the real problem: most of it is for reducing noise in the system and ignoring short random spikes. Weird that the human brain is still the best garbage filter ever designed.
I'm not even sure it is actually passing the full Turing test. From the article I noticed one of the researchers saying: "That's how we plan to pass this limited version of the Turing test." Anyone know what he means by this being a "limited" version of the Turing test?
Depends on what you are doing. If you are working on web apps and some Open Source projects you can get away with it. But start doing some digital signal processing/ encryption and you'll find out that functional analysis, abstract algebra, and the Fourier transform are important tools.
It's not that simple. If there is an SUV in front of my poor Carolla I can't see anything.
You make a valid point about stupid patents. I'm sure every engineer has seen patents that got them to roar with laughter at how obvious they are. But the fix isn't to have bribed senators use eminent domain and billion$ of tax dollars to swipe a company's property. The fix is to get the patent office to stop being so liberal in the granting of patents.
I doubt that what they are doing is even constitutional. The Constitution sets up clear property rights on patents. There is no escape clause anywhere in it that allows the Congress to pass laws nullifying a patent.
I understand what you're saying but military spending actually does get people clothed and fed. Look at what military spending has produced: televisions, HD-tv, computers, semiconductors, atomic energy, quantum physics, the internet, jet engines, GPS, etc. As contradictory as it may seem, most of the technology that makes people's lives better was actually originally designed for military purposes - non-military research yields far fewer fruits. I'm all for a smaller military but not for smaller military research budgets.
That's kind of my point. The market here is being filled domestically because it's a crappy career choice. When people were being paid well and treated well in the late 90's people were flocking to engineering. Now there are less and less engineering majors because people know they are much better off with a medical degree, law degree, or going into finance.
This is bullshit. If foreigners are so smart, why do they have to come to the US for jobs?
Why are people surprised noone wants to go into engineering in the US: stagnant wages, offshoring, age discrimination, long hours. It's a shitty way to waste $100k on an education.
I was starting to get the signs of carpal tunnel: I was losing feeling in my left thumb. For all of you younger engineers, losing sensation in your fingers is the first sign of early carpal tunnel syndrome.
I bought a Kinesis for about $300, and my hands healed within a couple weeks. It's a good 5 years later and I am still happy with my investment.
I do this all the time on decent sized projects. I am mostly working on control loops. For most things I will do informal proofs. For some very critical sections, I will go the full route and do a full proof. And yes, it does work. Spending an extra couple hours proving things is time well-spent when a full circuit compile will take me 2 hours. And yeah, if you use formal methods, except for syntax errors 90% of stuff works perfectly on the first run.
One section of code was taking up too much chip area. I used a formal method and cut the circuit size 80% with no loss in speed. So finally, formal methods do produce leaner and more bug-free code. More success stories from a simple google search http://shemesh.larc.nasa.gov/fm/fm-what.html.
I have a strong suspicion that you have never worked on an embedded system of any size.
There are a lot of heinous-looking things that pop up in embedded software that are absolutely necessary to get around hardware problems, timing issues, etc.
Try refactoring that stuff out and see how fast the whole system will crash into a pile of rubble.