Right, my point was that the computer analogy wasn't a good one. For precisely the reason(s) you describe.... this was the point of my "There's a lot of feedback" remark.
To use your analogy... bioinformatics research is proceeding rapidly down both the "bytecode" and "virtual machine" avenues of inquiry. Some people are actively working to uncover the structures of proteins and other relevant bio-molecules (see, for instance, this year's nobel prize in chemistry... kurt wuthrich pioneered the application of NMR to large molecules like proteins and dna, for the purpose of determining structure). Other people are trying to find the patterns in the dna itself that encode portions of the cell's regulatory apparatus. Other people are trying to characterize which genes are complicit in which pathways of the cell...
The analogy of "bytecode" and "virtual machine" is flawed anyway: it gives the impression that the cell's transcriptional apparatus is just an interpreter (highly parallelized, sure) that chugs down the DNA, reads the "code", and produces the appropriate proteins to do the cell's business. But that's misleading... for the most part, the cell's transcriptional activity is in some steady state, until outside stimuli signal it (in a complex way) to change one part of the humming machine, and then that change cascades to other portions of the cell's transcriptional activity, until the cell has reconfigured itself to handle the stimulus. There's a lot of feedback between the proteins and the dna (the transcriptional apparatus _is_ protein), etc.
A better analogy might be... well, I'm not sure there's a decent analogy at all. Maybe the "cell is a virtual machine," and outside stimuli are a form of programming language... Bleh, that's no good at all either.
At any rate, your post makes it seem like bioinformatics researchers have made a universal choice to put their research priorities in the wrong order... but that's certainly not the case. Working to decode the cell's apparatus in different ways simultaneously makes everyone's research more productive and useful.
Like I said before: it probably was unintentional in this case. My complaint is that it still would have been called a penalty, unintentional or not, in most other situations.
As for seeing or not seeing the handball: my impression about the ref running in was that it was due to the fact that, immediately after the handball, the ball fell into Kahn's hands and a US player slid into him. Most refs are going to run in to protect a goalkeeper, and most refs will act decisive (even if they're not sure about _everything_ that happened) to assert their authority and keep the game calm (I know this from experience: I've been a ref myself for several years). And I bet (neither one of us can say for sure) that if he felt at the time that there was controversy, it was due to the possibility that the ball crossed over the line... or at least the possibility that the US players would say it did (personally, I don't believe it did).
I know the ref (Hugh Dallas) was Scottish, and it has nothing to do with him paying attention or not. Even the best refs make mistakes, I'm not blaming him (and I'm not alleging a conspiracy). It wasn't an intentional mistake, just the type of call that typically goes against the US, that's all.
As for European refs: I thought the Swiss ref (Urs Maier?) was pretty bad in a couple of games. And the Portugese ref? What was his name?
Refereeing problems didn't skip the Europeans either.
I don't really believe that "intentional" is _necessarily_ tied to whether you move your arm at the ball or not. I mean, someone _could_ "intentionally" leave their arm in the path of the ball, knowing that the ball will hit it and fall to their side. In that situation, the hand and arm would be stationary, but the right call would (should) be a handball. The rules say something about "intent," but nothing about whether you have to move your hand or not.
I'm not saying that was the case here, with Frings on the goalline... but my complaint is that that kind of stuff is routinely called a handball anyway. The US would have scored was it not for that hand. More likely is that Hugh Dallas just didn't see the handball, and uses the "well, it probably was unintentional" defense after the fact.
The camera crews weren't American -- I actually thought that the camera angles were alright, at least we got a lot fewer crowd shots. Have you watched an NBA game recently? 30% of the game is crowd shots, or interviews with someone in the crowd, etc etc.
Also, I don't think that ABC/ESPN spent any money to "win" the rights to show this. They were basically bribed (or at least, the deal was sweetened _considerably_) by the same soccer millionaires who've been propping up the MLS.
Of course, the ratings were decent enough that they'll probably be willing to spend their own money on it in 2006.
I haven't read Wolfram's book, but you should try reading:
The Algorithmic Beauty of Plants by Przemyslaw Prusinkiewicz, Aristid Lindenmayer
before saying that he's hit on something completely new and original here. At the very least, you can check the book out of a library and look at the amazing pictures of lifelike plants, generated from these simple systems (L-Systems, named after the author "lindenmayer").
But it is on the web. Try here. I think that it should be generally accessible (i.e., I hope I'm not getting to it based solely on an institutional subscription).
Actually, I think your explanation was a bit confusing.... But that doesn't mean it's not right:-). I've become convinced of my mistake with help from the web, but thank you anyway for being patient and trying to explain it to me. Agh, need more coffee this early in the morning...
After reading about it in these comments, I went and found this solution online, here. This convinced me completely... I was drawing the same diagram as this guy "Peter", but I was making the same mistake he had made. Then I realized my mistake (by drawing out the possibilities as partitions of a unit square) and his second diagram confirms it.
It's never to early in the morning to feel dumb, I guess.
I heard him speak at MIT, and read a paper of his that was published in the Bulletin of the American Mathematical Society... On the Mathematical Foundations of Machine Learning, with Felipe Cucker I think. That was published in Oct. 2001, which qualifies as within the last 5 years, right?
Prize is behind door 1, you pick door 1: Monty opens either door 2 or 3, you switch and you lose.
Prize is behind door 2, you pick door 1: Monty opens door 3, you switch to door 2 and win.
Prize is behind door 3, you pick door 1: Monty opens door 2, you switch to door 3 and win.
So, my question now is: why are there only 3 cases and not 4? It seems to be that your case #1 should be split up into two cases:
1a) Monty picks door 2
1b) Monty picks door 3
Either way, you switch and lose... and now the ratio is 1:1, winning to losing, and there's no apparent advantage to switching.
Or maybe you'd like to break it down in the sense that the fundamental actions aren't which doors are picked but what type of doors are picked:
1. You pick right, and Monty shows you a wrong one. You switch and lose.
2. You pick wrong, and Monty shows you the other wrong one. You switch and win.
Now there are 2 cases instead of (my) 4 cases, and it's still a 1:1 thing: half the time you switch and win, half the time you switch and lose. Either way, there's no advantage to switching.
Now, I'm no mathematician. And I read the Marilyn article, but I've read a few probability theory books in the meantime too. And I'll admit that I could be wrong here, after all, I'm only human: But it just seems to me that you're not gaining any (any!) additional information from what Mr. Hall shows you. You know, before you even pick your door in the first place that you'll be shown a wrong door. At which point, there will be two doors left, one right, one wrong, and you will have chosen one of them. There still doesn't seem (to me) to be any advantage to switching.
Look, if you want to approach it from a classical viewpoint, with a set of elementary outcomes and a sigma field of events (subsets of those outcomes), and probabilities assigned to those events... Well, my intuition is that we'll get a situation more like the first one I described above (splitting case 1 into 1a and 1b), but I have to run to work. Ack. Maybe I can post it later in the day, if you're still interested in talking about it.
Um... I'm confused. What do you mean by "fair" and "unfair?"
The way I always heard the problem, it was what you were describing as unfair: the contestant picks a door, and monty hall shows him one of the wrong doors. And then you get a chance to switch. But I don't see why anyone would ever switch, if you know going in that you'll be shown a wrong door. Nothing has changed between the time when you chose a door in the first place, and the time when you can switch doors if you want.
Basically, my question is: why switch? There's no difference between the two remaining doors (in terms of your belief about what may be behind them) after Mr. Hall opens up a wrong door...
In other words, what I don't agree with is when you say that there's an "advantage to switching" when the show is unfair (as we both agree it is...)
Wasn't this a question on "Ask Marilyn" a while ago?
Hey, I don't mean to be argumentative, but I think you're misunderstanding what I'm saying...
So maybe saying that the Nyquist frequency has nothing to do with human hearing is a bit of a reach. But I stand by my point that there's no such thing as the "Nyquist frequency of human hearing."
Let me quickly quote my first post:
Probably the human ear would be able to pick up on aliasing if the sampling rate was too slow... and I'm certainly willing to believe that the sampling is so fine on CDs that it's indistinguishable to the normal human ear... But those are different issues.
In other words, we agree on one point: the sampling of the original signal has a lot to do with the quality of what's on the CD, but beyond a certain point increases in signal quality are indistiguishable to the human ear.
We agree on that. But I maintain that the "Nyquist frequency" is something that applies to signals, not to human ears. Given a discrete time series, it's the maximum frequency of a signal that can be reconstructed unambiguously from those discrete samples. (Equivalently, and this was the way I heard it in my a signal processing class I took a few years ago, it's the frequency at which the signal must at least be sampled at to perfectly reconstruct that signal).
Again, my point was that although sampling frequency has a lot to do with human ears and music, the "Nyquist frequency" is a number that is a property of the signal and its sample.
Um, the "Nyquist frequency" has nothing to do with "human hearing." It's a function of the signal (i.e., the signal which is sampled to get the discrete time series that is then encoded on the CD) itself. It has nothing to do with _who_ is listening to the music. If you sample a signal at a frequency lower than its Nyquist frequency, you get aliasing, that's all.
Probably the human ear would be able to pick up on aliasing if the sampling rate was too slow... and I'm certainly willing to believe that the sampling is so fine on CDs that it's indistinguishable to the normal human ear... But those are different issues.
I'm just frustrated with your use of the phrase "Nyquist frequency of human hearing." Get your math straight before you go calling other people "snobby."
> Work as one of those weird New York street vendors selling unidenfied meat on a stick.
And then wash it all down with crab juice...
mmmm.... crab juice....
Not to nit-pick, and I agree with what you're saying, but...
Just so you know: you can cast "aspersions" on things, but not "dispersions." Casting dispersions doesn't make sense... it's like saying "stop casting diffusions on the US."
Maybe you're thinking of "Chinese Room?" As I remember, the Chinese Room was an analogy used by a philosopher (Searle?) to make a point about AI and computers. That analogy was something along the lines of:
Imagine a room with with two slots in the walls. Inside the room sits a man with a massive instruction book. Outside the room, there are people who speak Chinese who play a Turing-test like game with the monolithic room: they write down questions (in Chinese) on pieces of paper, and insert them into the room through a slot. Inside the room, the man doesn't speak or read Chinese. But his book contains an infinitely large set of instructions: for each set of (unintelligible to him) symbols he sees on the input, he writes down another set of symbols on another paper and sticks it through the wall. The book was written by someone who understands Chinese, and contains an "intelligent" answer to every possible question he could be asked. The man doesn't know this, however... So the people outside the room submit questions, get answers, and conclude "The room has intelligence!" But inside, all that's going on is a rote following of instructions...
I think the point was that simple passing of the Turing test didn't mean intelligence; that humans have something a machine could never have (at least, a machine as we know them). There are responses, rebuttals, and counter-rebuttals of course...
It seems particularly applicable though to clean-room reverse engineering... except the engineer is what's in the box. Instead of having the book, the engineer gets the input and output... and it's his job to write the book.
"Plus, you lose the ability of quickly searching files that is available with most text editors."
Well, maybe, maybe not. The book comes with a CD that has (among other things) all the text from the book, along with line numbers. There's also a nice little emacs script to search the text...
Branch and Bound is an interesting technique, but it won't help you in chess. I work in computational biology, where some biological problems are simulated through recursive searching through trees (i.e., something like protein threading). I know people have used branch and bound for that problem, and it doesn't make exponential problems possible to solve (although it may help the running time for some things). The point is, it's hard to predict how branch and bound will affect your search.
For that matter, what is branch and bound? You're pruning elements of your search tree that are _provably_ worse than the optimal solution. I think you could do this in some cases in chess (i.e., you can prune trees which lead to loss immediately, perhaps) but I can't see how you'd implement it to be able to remove massive sections of your search tree, enough to make the problem feasible... that's the whole point of chess: sometimes a move looks bad, but you have to investigate (i.e., search) to find out for sure.
Branch and bound is interesting, but I don't think it's applicable here...
If you read the FAQ, it also refers to Gnutella as a "servent" in a "servent-to-servent" network, because any Gnutella can be either a client or a server for other clients, with no pre-arranged restrictions...
um, isn't that what a "peer" is? a node in the network that doesn't have a pre-arranged server or client status?
I feel like there's a bias sometimes on slashdot, towards a computer proficiency only useful in the computer industry. To that end, perhaps a completely-online university would have a place...
However, the meaning of "university" has always been (in my mind) tied up with the idea of research, and professor-graduate student associations that teach people how to do research, how to be part of academia (in the best sense of the word). This is the thing that would be lacking in an online institution. Academic research is an exercise that must be conducted in person; learning to be a researcher can't happen when you only know your professor through email and "streaming video". Research is a culture that grows in the daily interactions of a student and a teacher... and when it does occur, it produces ideas that often motivate or create entire industries (see the effect of Computer Science research on the "Computer Industry" for proof).
I'm a dartmouth student in the CS department here, and I can tell you first hand that not all Ivy schools are mediocre undergrad institutions. Dartmouth's main focus is, and always has been, undergrad educuation: graduate students here are an extreme minority, and not even present in all disciplines. Undergrad's regularly have chances for research with professors, and take classes from full professors (not just grad TA's), even starting as freshmen. It's even reflected in the name of the school: "Dartmouth College", not "Dartmouth University".
You might try reading the pages that were linked to: if you wanted to vote, you have to enter a PIN that was mailed out to before the election. I assume that, if someone tries to use an invalid PIN, or a PIN that isn't 'real,' then you can't vote.
I guess if you know someone else's PIN... but then again, that's not very different than knowing someone's password. Either way, there's not much you can do about it...
I feel like the assumption is that compilers can generate better code than a human programmer, not that they do. It seemed to me like the submission was kind of a mission statement for compilers/JIT-programs, a vision for the future of which Crusoe is perhaps a new iteration.
But clearly, a knowledgeable, good assembly coder can still beat a compiler some or most of the time.
I doubt that the idea of "tariffs" is one found only in Arkansas. If someone is conducting a business off of merchandise that they carry illegally through _US_ customs that's "fishy," no matter _how_ they operate in Asia.
Slashdot Mirror
Right, my point was that the computer analogy wasn't a good one. For precisely the reason(s) you describe.... this was the point of my "There's a lot of feedback" remark.
To use your analogy... bioinformatics research is proceeding rapidly down both the "bytecode" and "virtual machine" avenues of inquiry. Some people are actively working to uncover the structures of proteins and other relevant bio-molecules (see, for instance, this year's nobel prize in chemistry... kurt wuthrich pioneered the application of NMR to large molecules like proteins and dna, for the purpose of determining structure). Other people are trying to find the patterns in the dna itself that encode portions of the cell's regulatory apparatus. Other people are trying to characterize which genes are complicit in which pathways of the cell...
The analogy of "bytecode" and "virtual machine" is flawed anyway: it gives the impression that the cell's transcriptional apparatus is just an interpreter (highly parallelized, sure) that chugs down the DNA, reads the "code", and produces the appropriate proteins to do the cell's business. But that's misleading... for the most part, the cell's transcriptional activity is in some steady state, until outside stimuli signal it (in a complex way) to change one part of the humming machine, and then that change cascades to other portions of the cell's transcriptional activity, until the cell has reconfigured itself to handle the stimulus. There's a lot of feedback between the proteins and the dna (the transcriptional apparatus _is_ protein), etc.
A better analogy might be... well, I'm not sure there's a decent analogy at all. Maybe the "cell is a virtual machine," and outside stimuli are a form of programming language... Bleh, that's no good at all either.
At any rate, your post makes it seem like bioinformatics researchers have made a universal choice to put their research priorities in the wrong order... but that's certainly not the case. Working to decode the cell's apparatus in different ways simultaneously makes everyone's research more productive and useful.
Like I said before: it probably was unintentional in this case. My complaint is that it still would have been called a penalty, unintentional or not, in most other situations.
As for seeing or not seeing the handball: my impression about the ref running in was that it was due to the fact that, immediately after the handball, the ball fell into Kahn's hands and a US player slid into him. Most refs are going to run in to protect a goalkeeper, and most refs will act decisive (even if they're not sure about _everything_ that happened) to assert their authority and keep the game calm (I know this from experience: I've been a ref myself for several years). And I bet (neither one of us can say for sure) that if he felt at the time that there was controversy, it was due to the possibility that the ball crossed over the line... or at least the possibility that the US players would say it did (personally, I don't believe it did).
I know the ref (Hugh Dallas) was Scottish, and it has nothing to do with him paying attention or not. Even the best refs make mistakes, I'm not blaming him (and I'm not alleging a conspiracy). It wasn't an intentional mistake, just the type of call that typically goes against the US, that's all.
As for European refs: I thought the Swiss ref (Urs Maier?) was pretty bad in a couple of games. And the Portugese ref? What was his name?
Refereeing problems didn't skip the Europeans either.
I don't really believe that "intentional" is _necessarily_ tied to whether you move your arm at the ball or not. I mean, someone _could_ "intentionally" leave their arm in the path of the ball, knowing that the ball will hit it and fall to their side. In that situation, the hand and arm would be stationary, but the right call would (should) be a handball. The rules say something about "intent," but nothing about whether you have to move your hand or not.
I'm not saying that was the case here, with Frings on the goalline... but my complaint is that that kind of stuff is routinely called a handball anyway. The US would have scored was it not for that hand. More likely is that Hugh Dallas just didn't see the handball, and uses the "well, it probably was unintentional" defense after the fact.
The camera crews weren't American -- I actually thought that the camera angles were alright, at least we got a lot fewer crowd shots. Have you watched an NBA game recently? 30% of the game is crowd shots, or interviews with someone in the crowd, etc etc.
Also, I don't think that ABC/ESPN spent any money to "win" the rights to show this. They were basically bribed (or at least, the deal was sweetened _considerably_) by the same soccer millionaires who've been propping up the MLS.
Of course, the ratings were decent enough that they'll probably be willing to spend their own money on it in 2006.
I haven't read Wolfram's book, but you should try reading:
The Algorithmic Beauty of Plants
by Przemyslaw Prusinkiewicz, Aristid Lindenmayer
before saying that he's hit on something completely new and original here. At the very least, you can check the book out of a library and look at the amazing pictures of lifelike plants, generated from these simple systems (L-Systems, named after the author "lindenmayer").
True :-).
But it is on the web. Try here. I think that it should be generally accessible (i.e., I hope I'm not getting to it based solely on an institutional subscription).
Actually, I think your explanation was a bit confusing.... But that doesn't mean it's not right :-). I've become convinced of my mistake with help from the web, but thank you anyway for being patient and trying to explain it to me. Agh, need more coffee this early in the morning...
After reading about it in these comments, I went and found this solution online, here. This convinced me completely... I was drawing the same diagram as this guy "Peter", but I was making the same mistake he had made. Then I realized my mistake (by drawing out the possibilities as partitions of a unit square) and his second diagram confirms it.
It's never to early in the morning to feel dumb, I guess.
Do you mean the Steve Smale from Berkeley who won a Fields Medal?
smale bio
I heard him speak at MIT, and read a paper of his that was published in the Bulletin of the American Mathematical Society... On the Mathematical Foundations of Machine Learning, with Felipe Cucker I think. That was published in Oct. 2001, which qualifies as within the last 5 years, right?
Prize is behind door 1, you pick door 1: Monty opens either door 2 or 3, you switch and you lose.
Prize is behind door 2, you pick door 1: Monty opens door 3, you switch to door 2 and win.
Prize is behind door 3, you pick door 1: Monty opens door 2, you switch to door 3 and win.
So, my question now is: why are there only 3 cases and not 4? It seems to be that your case #1 should be split up into two cases:
1a) Monty picks door 2
1b) Monty picks door 3
Either way, you switch and lose... and now the ratio is 1:1, winning to losing, and there's no apparent advantage to switching.
Or maybe you'd like to break it down in the sense that the fundamental actions aren't which doors are picked but what type of doors are picked:
1. You pick right, and Monty shows you a wrong one. You switch and lose.
2. You pick wrong, and Monty shows you the other wrong one. You switch and win.
Now there are 2 cases instead of (my) 4 cases, and it's still a 1:1 thing: half the time you switch and win, half the time you switch and lose. Either way, there's no advantage to switching.
Now, I'm no mathematician. And I read the Marilyn article, but I've read a few probability theory books in the meantime too. And I'll admit that I could be wrong here, after all, I'm only human: But it just seems to me that you're not gaining any (any!) additional information from what Mr. Hall shows you. You know, before you even pick your door in the first place that you'll be shown a wrong door. At which point, there will be two doors left, one right, one wrong, and you will have chosen one of them. There still doesn't seem (to me) to be any advantage to switching.
Look, if you want to approach it from a classical viewpoint, with a set of elementary outcomes and a sigma field of events (subsets of those outcomes), and probabilities assigned to those events... Well, my intuition is that we'll get a situation more like the first one I described above (splitting case 1 into 1a and 1b), but I have to run to work. Ack. Maybe I can post it later in the day, if you're still interested in talking about it.
Um... I'm confused. What do you mean by "fair" and "unfair?"
The way I always heard the problem, it was what you were describing as unfair: the contestant picks a door, and monty hall shows him one of the wrong doors. And then you get a chance to switch. But I don't see why anyone would ever switch, if you know going in that you'll be shown a wrong door. Nothing has changed between the time when you chose a door in the first place, and the time when you can switch doors if you want.
Basically, my question is: why switch? There's no difference between the two remaining doors (in terms of your belief about what may be behind them) after Mr. Hall opens up a wrong door...
In other words, what I don't agree with is when you say that there's an "advantage to switching" when the show is unfair (as we both agree it is...)
Wasn't this a question on "Ask Marilyn" a while ago?
Hey, I don't mean to be argumentative, but I think you're misunderstanding what I'm saying...
So maybe saying that the Nyquist frequency has nothing to do with human hearing is a bit of a reach. But I stand by my point that there's no such thing as the "Nyquist frequency of human hearing."
Let me quickly quote my first post:
Probably the human ear would be able to pick up on aliasing if the sampling rate was too slow... and I'm certainly willing to believe that the sampling is so fine on CDs that it's indistinguishable to the normal human ear... But those are different issues.
In other words, we agree on one point: the sampling of the original signal has a lot to do with the quality of what's on the CD, but beyond a certain point increases in signal quality are indistiguishable to the human ear.
We agree on that. But I maintain that the "Nyquist frequency" is something that applies to signals, not to human ears. Given a discrete time series, it's the maximum frequency of a signal that can be reconstructed unambiguously from those discrete samples. (Equivalently, and this was the way I heard it in my a signal processing class I took a few years ago, it's the frequency at which the signal must at least be sampled at to perfectly reconstruct that signal).
Again, my point was that although sampling frequency has a lot to do with human ears and music, the "Nyquist frequency" is a number that is a property of the signal and its sample.
Um, the "Nyquist frequency" has nothing to do with "human hearing." It's a function of the signal (i.e., the signal which is sampled to get the discrete time series that is then encoded on the CD) itself. It has nothing to do with _who_ is listening to the music. If you sample a signal at a frequency lower than its Nyquist frequency, you get aliasing, that's all.
Probably the human ear would be able to pick up on aliasing if the sampling rate was too slow... and I'm certainly willing to believe that the sampling is so fine on CDs that it's indistinguishable to the normal human ear... But those are different issues.
I'm just frustrated with your use of the phrase "Nyquist frequency of human hearing." Get your math straight before you go calling other people "snobby."
> Work as one of those weird New York street vendors selling unidenfied meat on a stick.
And then wash it all down with crab juice...
mmmm.... crab juice....
Not to nit-pick, and I agree with what you're saying, but...
Just so you know: you can cast "aspersions" on things, but not "dispersions." Casting dispersions doesn't make sense... it's like saying "stop casting diffusions on the US."
Again, sorry for being so anal.
Maybe you're thinking of "Chinese Room?" As I remember, the Chinese Room was an analogy used by a philosopher (Searle?) to make a point about AI and computers. That analogy was something along the lines of:
Imagine a room with with two slots in the walls. Inside the room sits a man with a massive instruction book. Outside the room, there are people who speak Chinese who play a Turing-test like game with the monolithic room: they write down questions (in Chinese) on pieces of paper, and insert them into the room through a slot. Inside the room, the man doesn't speak or read Chinese. But his book contains an infinitely large set of instructions: for each set of (unintelligible to him) symbols he sees on the input, he writes down another set of symbols on another paper and sticks it through the wall. The book was written by someone who understands Chinese, and contains an "intelligent" answer to every possible question he could be asked. The man doesn't know this, however... So the people outside the room submit questions, get answers, and conclude "The room has intelligence!" But inside, all that's going on is a rote following of instructions...
I think the point was that simple passing of the Turing test didn't mean intelligence; that humans have something a machine could never have (at least, a machine as we know them). There are responses, rebuttals, and counter-rebuttals of course...
It seems particularly applicable though to clean-room reverse engineering... except the engineer is what's in the box. Instead of having the book, the engineer gets the input and output... and it's his job to write the book.
Actually, I think the story says that the CEO of Connectix (not the court) touted this as indicative of the law's support for 'broad consumer choice.'
"Plus, you lose the ability of quickly searching files that is available with most text editors."
Well, maybe, maybe not. The book comes with a CD that has (among other things) all the text from the book, along with line numbers. There's also a nice little emacs script to search the text...
Branch and Bound is an interesting technique, but it won't help you in chess. I work in computational biology, where some biological problems are simulated through recursive searching through trees (i.e., something like protein threading). I know people have used branch and bound for that problem, and it doesn't make exponential problems possible to solve (although it may help the running time for some things). The point is, it's hard to predict how branch and bound will affect your search.
For that matter, what is branch and bound? You're pruning elements of your search tree that are _provably_ worse than the optimal solution. I think you could do this in some cases in chess (i.e., you can prune trees which lead to loss immediately, perhaps) but I can't see how you'd implement it to be able to remove massive sections of your search tree, enough to make the problem feasible... that's the whole point of chess: sometimes a move looks bad, but you have to investigate (i.e., search) to find out for sure.
Branch and bound is interesting, but I don't think it's applicable here...
If you read the FAQ, it also refers to Gnutella as a "servent" in a "servent-to-servent" network, because any Gnutella can be either a client or a server for other clients, with no pre-arranged restrictions...
um, isn't that what a "peer" is? a node in the network that doesn't have a pre-arranged server or client status?
I feel like there's a bias sometimes on slashdot, towards a computer proficiency only useful in the computer industry. To that end, perhaps a completely-online university would have a place...
However, the meaning of "university" has always been (in my mind) tied up with the idea of research, and professor-graduate student associations that teach people how to do research, how to be part of academia (in the best sense of the word). This is the thing that would be lacking in an online institution. Academic research is an exercise that must be conducted in person; learning to be a researcher can't happen when you only know your professor through email and "streaming video". Research is a culture that grows in the daily interactions of a student and a teacher... and when it does occur, it produces ideas that often motivate or create entire industries (see the effect of Computer Science research on the "Computer Industry" for proof).
I'm a dartmouth student in the CS department here, and I can tell you first hand that not all Ivy schools are mediocre undergrad institutions. Dartmouth's main focus is, and always has been, undergrad educuation: graduate students here are an extreme minority, and not even present in all disciplines. Undergrad's regularly have chances for research with professors, and take classes from full professors (not just grad TA's), even starting as freshmen. It's even reflected in the name of the school: "Dartmouth College", not "Dartmouth University".
You might try reading the pages that were linked to: if you wanted to vote, you have to enter a PIN that was mailed out to before the election. I assume that, if someone tries to use an invalid PIN, or a PIN that isn't 'real,' then you can't vote.
I guess if you know someone else's PIN... but then again, that's not very different than knowing someone's password. Either way, there's not much you can do about it...
I feel like the assumption is that compilers can generate better code than a human programmer, not that they do. It seemed to me like the submission was kind of a mission statement for compilers/JIT-programs, a vision for the future of which Crusoe is perhaps a new iteration.
But clearly, a knowledgeable, good assembly coder can still beat a compiler some or most of the time.
I doubt that the idea of "tariffs" is one found only in Arkansas. If someone is conducting a business off of merchandise that they carry illegally through _US_ customs that's "fishy," no matter _how_ they operate in Asia.