[small businesses] also employ 51 percent of private-sector workers
This must be with a very loose definition of "small business".
Businesses with 1-9 employees employ just 12% of the private workforce. Businesses with 1-49 employees employ just 49 percent of private sector employees. Moremover, this is in some ways an overcount: it counts "subcontractors" as small businesses even if they're essentially just a worker who's not being paid benefits. Also, franchises like McDonalds, Wendy's, NAPA auto parts, Super Cuts, etc., would fall under these categories as well, even though most people wouldn't think of them as a mom + pop "small business".
My source for these statistics: http://www.bls.gov/oes/2003/may/employment.pdf
Another thing about these numbers is that they go down every year... every year, "small business" is a smaller sector of the economy by employment.
The program effectively turns anyone's personal computer into a proxy server. Once the software is installed on a computer in, say, Canada, that person creates a contact list of trusted friends or family members in censored countries and sends his or her IP address to them. No advertising needed.
Man, why not just use tor? You would get good anonymity, AND the tor network could route blocked domains. Is the problem that they would start blocking all tor nodes??
given the state of things right now, it seems VoIP has a chance to become the only secure way to talk to someone over distances. If people can use an open source encryption scheme for their VoIP, the NSA will have significantly more trouble butting in on your conversations---even with the help of AT&T.
Blaming artists for their contract terms is like blaming a fast food restaurant worker for the low minimum wage. In the artist-label economic relationship, the labels have all the power. Every artist needs a label, no label needs any particular artist. A rich and famous artist is a horse of a different color, of course, but by that point, you probably owe a label a bunch of money (for all they "invested" in you).
The solution to the problem you're talking about would be some sort of collective bargaining unit for the artists, but this seems unlikely to happen given the diversity of the group, and the absence of any collective workplace per se.
You seem to be trying to give the impression that news organizations and businesses are different things. Most modern news organizations (NYTimes, CNN, Fox, WSJ, etc., etc.) are businesses. For them, the bottom line is the driving force. To the extent that being good at informing people is good for their bottom line, they try to be good at informing people. We are all familiar with the result, and capable of making our own judgements of how well this all works out.
Neither teachers nor students choose books. State boards choose books. Lies My Teacher Told Me is an excellent book for people interested in the world of textbook politics. (The book focuses on American History textbooks, but many of its points apply to others: biology, etc.)
He understood this situation; he himself had had to borrow to pay for all the research papers he read. (10% of those fees went to the researchers who wrote the papers; since Dan aimed for an academic career, he could hope that his own research papers, if frequently referenced, would bring in enough to repay this loan.)
Since when do royalties get paid for Academic papers? They don't. In fact, scientists/mathematicians also volunteer their time to peer-review articles that appear in journals... did you think journals paid to get articles reviewed? They don't. They assume the copyright, and print copies of the journal which they sell to institutions for hundreds of dollars. They don't even really do much typesetting anymore, thanks to LATEX. Even before the takeover of DRM, the crisis has already begun---simply because profits are the driving force between anything run by a business. And, like it or not, it is not always true that profits=progress.
What makes this even more significant is that it has been proven that if we find a polynomial solution for one NP problem, we can create solutions for any NP problem.
You mean: if we find a polynomial solution for any NP-complete problem. There are many NP problems with polynomial algorithms. (This is exactly the set P).
Another point which might illustrate how abstract this problem is:
You say playing chess (i.e., knowing the best move for any configuration) is in EXPTIME. Well, Chess is in P as well, technically speaking. In fact, playing chess requires only a constant time algorithm: there are a finite number of configurations on the board, and a list could "theoretically" be made with all the best moves in each situation. (In other words, such a list "exists", in the mathematical sense). Thus an algorithm which looks at the list and makes the best move runs in constant time in terms of the input size (which itself is bounded!) I haven't given the algorithm, but I've proven it exists, which is all that's required.
The result you are citing is that a generalized form a chess played on an n x n board (instead of 8 x 8) is in exptime, in terms of n.
This illustrates a point: the complexity of problems only makes sense ASYMPTOTICALLY. For problems of a fixed (or bounded) input size, the complexity class is meaningless (or, at the very least, uninteresting, as this example illustrates). The complexity of problems is almost always of strictly theoretical interest. For some problems (linear programming comes to mind), the algorithm used in practice is known to technically be exponential, but almost always works better than the theoretically best (e.g., polynomial time) algorithm. This illustrates another point: the complexity class is determined by the WORST CASE running time of algorithms, which is not always what we'd care about in real life.
I think people need an informal definition of NP. Here it goes:
A decision problem is a yes/no question. A decision problem is in P if we can solve it in polynomial time. Example: is n divisble by 3? The length of the input to the algorithm is log n (the number of binary digits in n). We can divide a number by 3 in quadratic time in the length of the input. So we can certainly decide in essentially (log n)^2 steps if n is divisible by 3. (log n)^2 is a polynomial (of degree 2) in the length of the input to the algorithm (log n). So this decision problem in in P.
Consider another decision problem: is n composite? It's not obvious how to quickly determine if a large number is composite, short of trying all possible divisors up to the square root of n. The square root of n is exponential in the length of the input (log n) so this does not give a polynomial time algorithm. On the other hand, if someones else knows the factorization of n, they can tell you the factorization, and you can quickly use it to check that n is composite (by multiplying the factors together). THIS means that this decision problem is NP: if the answer is yes, there is a (short) "witness" that, if someone tells you the witness, lets you check in polynomial time that the answer is yes.
It turns out that this second problem is actually in P as well--that is, you can solve this problem without the witness. This was proved in 2002, and is not simple at all.
If P=NP, then this would always be true: anything you can solve quickly with the right witness can be solved without the witness. Proving P!=NP amounts to proving that there is some problem in NP which cannot be solved in polynomial time. There are many problems in NP known to be at least as hard as all the rest (called NP-complete), so if P!=NP, these can't be solved in polynomial time. Now, you just need to prove this...
It is completely wrong to say that NP problems are those that can't be solved in polynomial time, there are plenty of of problems that are non-polynomial and way harder than NP problems (playing chess is a good example, that belongs to complexity class EXPTIME).
Moreover, EVERY problem decideable in polynomial time is in NP. P is a SUBSET of NP.
I hate to say it, but NP stands for nondeterministc polynomial time, and not non-polynomial. This doesn't really make sense if your used to the algorithm/witness definition of NP, but NP is also defined as the class of languages decidable by nondeterministic Turing machines in polynomial time.
NP couldn't mean non-polynomial. First of all, of course P!=NP is still a conjecture. Also, P is a SUBSET of NP. That is, every polynomial solvable language is in NP.
I remember when I was in high school and someone first explained the P=NP problem to me. This was certainly someone who was very smart. I remember he had made big bucks at Microsoft doing some sort of software work. He told me he was reading a book about the problem (I'm not sure which one, there are many), and was going to "work on it". He told me about the millenium prize competition. But he said something else that really underlined for me the disconnect between Academia and the business world:
He told me that if he he solved the problem by showing P=NP (instead of P!=NP, which "most mathematicians believe"), he wouldn't publish his proof. Instead, he would setup a website that would take credit card payments to solve problems quickly (for example, packing boxes into the back of a UPS truck, or various traveling salesman problems). At the time, I though this was a little antisocial, but not much more.
Later, when I had more mathematical training, I looked back on this and realized how revealing this attitude was: of course, if someone proves P=NP, the proof will almost certainly not be accompanied by practical algorithms which are significantly better than those used already for problems on most scales. Of course, the idea that he was going to solve this problem without any collaboration or formal education in logic or complexity theory demonstrated the arrogance typical of many super-successful business-people. I can't help but remark that for all the stupid patents on software "ideas" and sometimes algorithms, we're lucky that, most of the time, theoretical advances are made not by people like this... and and so people publish their results, and are rewarded with respect rather than dollars.
Imagine the state of our theoretical knowledge in mathematics and computer science if, even in Academia, every discovery of a new algorithm or idea resulted in a patent application, and was jealously guarded as a secret which could produce profit. Unfortunately, this is already largely the state of things in the wet sciences (unnecessarily so, I would argue, and point to mathematics as my evidence).
As for the wiki thing: I don't think most ordinary people are like this guy, so hey, good for the wiki. (I think this attitude is taught by the business world, and not somehow the other way around). Unfortunately, I fear that the millenium problems are deep enough that amateurs will have trouble making a big impact. There are a few amateur contributions to mathematics occasionally, but there hasn't been a significant one in a long time. (The last was arguably by Marjorie Rice, a housewife who essentially resolved the question of the number of different ways to tile the plane with convex pentagons). Astronomy is probably the last big field where amateurs play a really significant role.
Not to say that makes him wrong, but something to keep in mind. Here is his wikipedia article. (Although, by the time you read this, it will probably have been changed by this post's childerens' authors....)
This begs a logical question: If hardware compatability is so difficult and expensive for Windows, why doesn't Microsoft just standardize it? Write a single (or at most, a few) modem/network card/whatever drivers, that manufacturers have to make the hardware conform to. I've never understood why this is technically a difficult or impractical solution. Can somebody explain this? Why do we need a hundred modem drivers? Is there really no other way?
In lieue of a technical explanation, I've assumed that this could be considered good for microsoft: making hardware compatiability an expensive task is good for the market giant, I guess (as the present discussion indicates).
didn't like lindows at all... they reviewed the Walmart Lindows pcs. If you have a CR subscription (or know someone who does) it's a good article to read from the standpoint of understanding what still stands in the way of mainstream acceptance of these distros.
Slashdot Mirror
Are IT people more mobile than truckers?
How long until they outlaw spam blockers which don't give "legitimate marketers" a backdoor?
given the state of things right now, it seems VoIP has a chance to become the only secure way to talk to someone over distances. If people can use an open source encryption scheme for their VoIP, the NSA will have significantly more trouble butting in on your conversations---even with the help of AT&T.
Blaming artists for their contract terms is like blaming a fast food restaurant worker for the low minimum wage. In the artist-label economic relationship, the labels have all the power. Every artist needs a label, no label needs any particular artist. A rich and famous artist is a horse of a different color, of course, but by that point, you probably owe a label a bunch of money (for all they "invested" in you). The solution to the problem you're talking about would be some sort of collective bargaining unit for the artists, but this seems unlikely to happen given the diversity of the group, and the absence of any collective workplace per se.
You seem to be trying to give the impression that news organizations and businesses are different things. Most modern news organizations (NYTimes, CNN, Fox, WSJ, etc., etc.) are businesses. For them, the bottom line is the driving force. To the extent that being good at informing people is good for their bottom line, they try to be good at informing people. We are all familiar with the result, and capable of making our own judgements of how well this all works out.
I guess I should clarify: I'm discussing selection of k-12 textbooks, even though TFA is primarily about college textbooks.
Neither teachers nor students choose books. State boards choose books. Lies My Teacher Told Me is an excellent book for people interested in the world of textbook politics. (The book focuses on American History textbooks, but many of its points apply to others: biology, etc.)
It seems like this law is more about distorting/destroying/etc an actual physical piece of artwork---not making another similar but "distorted' piece.
Another point which might illustrate how abstract this problem is: You say playing chess (i.e., knowing the best move for any configuration) is in EXPTIME. Well, Chess is in P as well, technically speaking. In fact, playing chess requires only a constant time algorithm: there are a finite number of configurations on the board, and a list could "theoretically" be made with all the best moves in each situation. (In other words, such a list "exists", in the mathematical sense). Thus an algorithm which looks at the list and makes the best move runs in constant time in terms of the input size (which itself is bounded!) I haven't given the algorithm, but I've proven it exists, which is all that's required. The result you are citing is that a generalized form a chess played on an n x n board (instead of 8 x 8) is in exptime, in terms of n. This illustrates a point: the complexity of problems only makes sense ASYMPTOTICALLY. For problems of a fixed (or bounded) input size, the complexity class is meaningless (or, at the very least, uninteresting, as this example illustrates). The complexity of problems is almost always of strictly theoretical interest. For some problems (linear programming comes to mind), the algorithm used in practice is known to technically be exponential, but almost always works better than the theoretically best (e.g., polynomial time) algorithm. This illustrates another point: the complexity class is determined by the WORST CASE running time of algorithms, which is not always what we'd care about in real life.
I think people need an informal definition of NP. Here it goes: A decision problem is a yes/no question. A decision problem is in P if we can solve it in polynomial time. Example: is n divisble by 3? The length of the input to the algorithm is log n (the number of binary digits in n). We can divide a number by 3 in quadratic time in the length of the input. So we can certainly decide in essentially (log n)^2 steps if n is divisible by 3. (log n)^2 is a polynomial (of degree 2) in the length of the input to the algorithm (log n). So this decision problem in in P. Consider another decision problem: is n composite? It's not obvious how to quickly determine if a large number is composite, short of trying all possible divisors up to the square root of n. The square root of n is exponential in the length of the input (log n) so this does not give a polynomial time algorithm. On the other hand, if someones else knows the factorization of n, they can tell you the factorization, and you can quickly use it to check that n is composite (by multiplying the factors together). THIS means that this decision problem is NP: if the answer is yes, there is a (short) "witness" that, if someone tells you the witness, lets you check in polynomial time that the answer is yes. It turns out that this second problem is actually in P as well--that is, you can solve this problem without the witness. This was proved in 2002, and is not simple at all. If P=NP, then this would always be true: anything you can solve quickly with the right witness can be solved without the witness. Proving P!=NP amounts to proving that there is some problem in NP which cannot be solved in polynomial time. There are many problems in NP known to be at least as hard as all the rest (called NP-complete), so if P!=NP, these can't be solved in polynomial time. Now, you just need to prove this...
I hate to say it, but NP stands for nondeterministc polynomial time, and not non-polynomial. This doesn't really make sense if your used to the algorithm/witness definition of NP, but NP is also defined as the class of languages decidable by nondeterministic Turing machines in polynomial time.
NP couldn't mean non-polynomial. First of all, of course P!=NP is still a conjecture. Also, P is a SUBSET of NP. That is, every polynomial solvable language is in NP.
I remember when I was in high school and someone first explained the P=NP problem to me. This was certainly someone who was very smart. I remember he had made big bucks at Microsoft doing some sort of software work. He told me he was reading a book about the problem (I'm not sure which one, there are many), and was going to "work on it". He told me about the millenium prize competition. But he said something else that really underlined for me the disconnect between Academia and the business world:
He told me that if he he solved the problem by showing P=NP (instead of P!=NP, which "most mathematicians believe"), he wouldn't publish his proof. Instead, he would setup a website that would take credit card payments to solve problems quickly (for example, packing boxes into the back of a UPS truck, or various traveling salesman problems). At the time, I though this was a little antisocial, but not much more.
Later, when I had more mathematical training, I looked back on this and realized how revealing this attitude was: of course, if someone proves P=NP, the proof will almost certainly not be accompanied by practical algorithms which are significantly better than those used already for problems on most scales. Of course, the idea that he was going to solve this problem without any collaboration or formal education in logic or complexity theory demonstrated the arrogance typical of many super-successful business-people. I can't help but remark that for all the stupid patents on software "ideas" and sometimes algorithms, we're lucky that, most of the time, theoretical advances are made not by people like this... and and so people publish their results, and are rewarded with respect rather than dollars.
Imagine the state of our theoretical knowledge in mathematics and computer science if, even in Academia, every discovery of a new algorithm or idea resulted in a patent application, and was jealously guarded as a secret which could produce profit. Unfortunately, this is already largely the state of things in the wet sciences (unnecessarily so, I would argue, and point to mathematics as my evidence).
As for the wiki thing: I don't think most ordinary people are like this guy, so hey, good for the wiki. (I think this attitude is taught by the business world, and not somehow the other way around). Unfortunately, I fear that the millenium problems are deep enough that amateurs will have trouble making a big impact. There are a few amateur contributions to mathematics occasionally, but there hasn't been a significant one in a long time. (The last was arguably by Marjorie Rice, a housewife who essentially resolved the question of the number of different ways to tile the plane with convex pentagons). Astronomy is probably the last big field where amateurs play a really significant role.
Not to say that makes him wrong, but something to keep in mind. Here is his wikipedia article. (Although, by the time you read this, it will probably have been changed by this post's childerens' authors....)
But you're all going to have to stop using google now. Stem cells are sacred---even metaphorical ones.
This begs a logical question:
If hardware compatability is so difficult and expensive for Windows, why doesn't Microsoft just standardize it? Write a single (or at most, a few) modem/network card/whatever drivers, that manufacturers have to make the hardware conform to. I've never understood why this is technically a difficult or impractical solution. Can somebody explain this? Why do we need a hundred modem drivers? Is there really no other way?
In lieue of a technical explanation, I've assumed that this could be considered good for microsoft: making hardware compatiability an expensive task is good for the market giant, I guess (as the present discussion indicates).
didn't like lindows at all... they reviewed the Walmart Lindows pcs. If you have a CR subscription (or know someone who does) it's a good article to read from the standpoint of understanding what still stands in the way of mainstream acceptance of these distros.
Headline: Kinderstart revenue jumps 200% as story reaches slashdot