Red is ~700 nm and violet is ~400 nm. A typical human can see light from the range of 390-750 nm with the aid of three cones. The three cones are the "red" cone (optimal at 564 nm), the "green" cone (optimal at 534 nm), and the "blue" cone (optimal at 420 nm).
Yes, I have noticed that some people post such original thoughts quite frequently. If I thought he'd win, I'd love to see him (Gore) run. However, I don't think he'll win, so I think that if he ran it'd just detract from his environmental message. As for Kerry, I didn't like him in '04 (although I preferred him to Bush), and I don't want to see him run in '08. Not sure who I do want in '08, but that's how it freakin' should be in 2006!
Of course, many do not. Probably more do not. However, Al Gore (the anti-enviros favorite whipping boy) does. That was one of the things I really liked about him in 2000. He understood technology and respected the environment. (Not that I want him to run in '08. I think it would detract from his current campaign.)
I hope I am wrong, but am willing to bet we won't have anything except the ISS (if we have even that) by 2020. The only possible exception might be if the Chinese put up something similar to ISS... but even that will be a far cry from anything we are talking about today (or twenty years ago).
That's not much of a prediction (although as someone else pointed out Bigelow might prove you wrong). Currently, NASA's plan is for a lunar base in 2024. Therefore, even an optimist shouldn't expect one before then. A realist might guess 2030, and a pessimist might guess not in this century. Of course, by definition, I'm a realist.:)
I'm really happy that most processors are now 64 bit. 64 bit means it's easier to access more RAM. More RAM means I can simulate more neurons and synapses in less time.
You could have at least used scare quotes around that last "impeached". I'm definitely not in favor of changing the word impeached to mean what the average American thinks it means. As you said, Clinton was impeached. He just wasn't removed from office.
The better I have learned how to write scripts in bash, the easier it has become for me to quickly write a "script" at the command line. I know, you can do this with Python, etc., but for some jobs it's just easier and quicker to use bash. The more tools you have, the quicker you can solve your problems.
Also, I believe that many default distros don't have Python in them. So, if you're on a strange computer you can't count on it being there. Perl or bash, however, you'll almost always find. (This is the same reason I've tried to improve my vi skills even though I'm an emacs fan. I know, I know. Yes, I said emacs.)
I've heard that even though it is calving a lot of icebergs these days, it's getting enough snowfall that the total ice in the antartic is actually increasing. It's just increasing in a different place than the icebergs are coming from.
Also, it wasn't scientists who were talking about global cooling in the 70's. It was some of the same media types who now think that "both" sides of the global warming "debate" need to be discussed. I.e., journalists who didn't understand science, but want to sell subscriptions. I challenge you to find a single peer-reviewed article supporting global cooling in the 70's. In fact, you'll find that scientists in the 70's were already warning about global warming.
Actually, the volume of water decreases up to about 4 degrees C, IIRC. After that it begins increasing again, like any well-behaved liquid should. Not that this changes your central argument in any way.
Especially when what the report has probably lowered is only the worst case scenario. Hopefully, the worst case scenario will always be lessened as one approaches the date in question.
and B) I hope you're familiar with the concept of over-generalization? I.e., you can't extract very much meaningfully predictive information from a single example.
Normally, I'd be happy to argue the actual politics with you (rather than just the logic), but I know it would be fruitless (and that's not intended to be an insult in any way - it's admittedly fruitless in both directions), and I'm a little tired.
Not that I really expect the logic/math argument to be that much more successful...:)
I said we should argue the logic (i.e., the lack of meaning in the denominators) rather than the politics since we definitely won't agree on the politics (at least not completely). If there are X terrorists of type A, and Y terrorists of type B, then the proper sampling would be X/Y regardless of how many total people there are of types A and B - unless you have additional information that we didn't even discuss.
Here's the basic logic: let's assume you have the accurate P(T|A), where T=person is terrorist, and A=person is of type A, and P(T|B). In this case, we'll say (for sake of argument) that P(T|A) = 25/100,000,000 and P(T|B) = 25/1,000,000. I.e., P(T|A) = 0.01 * P(T|B). Now, let's assume that the probability of flying on an airplane is identical for groups A & B (i.e., I'm assuming no "a priori" information). That means that P(F|A) = P(F|B), where F=person is flying on a plane. Let's also assume (although I won't argue it) that P(F|A) is independent of P(T|A). This assumption of equality will mean that we won't actually need to use this probability.
Now, the question you're really trying to ask, when you're (sort of) asking what is the ratio of A to B that should be sampled is what is P(A|T) relative to P(B|T)? Now, Bayes' theorem tells us that P(A|T)P(T) = P(A,T) = P(T|A)P(A). Solving this for P(T) gives us P(T) = P(T|A)P(A)/P(A|T). By similar logic, P(T) = P(T|B)P(B)/P(B|T), so P(T|A)P(A)/P(A|T) = P(T|B)P(B)/P(B|T). Now, if we assume that the ratio of P(A)/P(B) = 100 (i.e., there are 100 times more A than B), and given our initial assumption that P(T|A)/P(T|B) = 0.01, we have that P(A|T) = P(B|T). I.e., for every A you search, you should search 1 (and not 100) B. Note: this means that you will be effectively sampling B at 100 times the rate of A.
Finally, factor those numbers with number of the particular profiles.
Let's say the Whitie number is 20. (I'm being conservative) Now divide that number by the number of white people who fit the profile of the Whities. That's 20/100,000,000. Now let's take the number of Muzzies and divide them by the muslim male population or 25/1,000,000. This means that a muslim male is more than 100 times more likely to commit a terrorist act than a white male.
So, I'll agree with your assessment if you'll agree that for every Whitie that is searched before boarding a plane, 100 Muzzies must be searched before boarding a plane. Fair enough for you?
Just to argue the logic and not the politics (because I'm certain neither of us will budge on that), your logic is flawed. First of all, that 25 should be 19 (unless I'm missing 6 from somewhere, which I'll admit is possible). So, we'll just let both numbers be 19. (If I really wanted to, I could come up with scores of "Columbinish" events, including one that happened at a school where I used to teach, but I won't go there.) As I'll try to explain, the denominator in the first paragraph is irrelevant for your profile rate in the second paragraph. If there are 100 times more "Whities" than "Muzzies" (gak!), then if you search one W for each one (not 100) M, then you'll be targeting them 100 times as much (since there are 1/100th as many to search). If you searched 100 M's for every 1 W (as you suggest), then you are targeting them 10,000 times as much despite your assertion that they are only 100 times more likely to be a terrorist.
Then, of course, there's the problem of identifying M's. Are we going to use crescent armbands or something? (Does that count towards Godwin?).
Yes, but how many of the Oklahoma City terrorists were Muslims? How many of the abortion clinic bombing terrorists were Muslims? How many of the Columbine terrorists were Muslims?
I can't hear it. If I turn my speakers up, I can hear the distortion in my speakers, but I'm not really hearing it. I know, because when I turn them down, I don't hear anything, but one of my colleagues can, and yells at me to turn it off.:D Hm, let me try that again. Yep, still works.:D (I only play it for a second or two. I'm not that mean.) Actually, from about 20-30 there's a significant drop-off in who can hear it, so don't be surprised if you're not really hearing it. Get a similarly aged (or younger) friend to help you verify whether or not you're hearing it.
And yes, there are people over 60 who can hear it, too. Very few, however.
Actually, I define "passing on the right" the same way you do, but I define "passing someone to your right" to mean that the person you are passing is to your right. So, yes, we're basically agreeing.
If it's a choice between a loudspeaker saying "you guys need to leave here" and this, well, then I'd rather have the loudspeaker.
There is a middle ground - you could always have the loudspeaker play this. (If you can't hear this, then you're probably over 30. I'm 36, and I can't hear it. It annoys the @$#! out of those who can hear it, though. I have it bookmarked.:D )
There's no doubt that different regions have different types of drivers, and I agree that if you're not passing someone you should get over, but how exactly do you define "substantially faster"? IMO, if you're passing someone to your right (and you're driving at least the speed limit), you need not make any apologies for not going "fast enough" for the people behind you. Sometimes this is complicated somewhat by hills and tractor trailers, of course. (I live near the Blue Ridge Parkway, where this might be a bigger issue than in Massachusetts.)
Also, I should confess that these days I only get on the highway about three to five times a year. I used to commute ~20 miles to work each way when I lived in Atlanta (~5 years ago), but now I commute <1 mile to work each way (via foot) now that I live in glorious Charlottesville.
But seriously, I'm fairly certain they are highly correlated. I drive the speed limit, and I'm always in the correct lane. It is very, very rare that I have had to pass someone going slower than the speed limit, but when I have - they are in the correct lane probably close to 99% of the time.
If you think otherwise, you're probably suffering from confirmation bias.
Slashdot Mirror
Red is ~700 nm and violet is ~400 nm. A typical human can see light from the range of 390-750 nm with the aid of three cones. The three cones are the "red" cone (optimal at 564 nm), the "green" cone (optimal at 534 nm), and the "blue" cone (optimal at 420 nm).
Yes, I have noticed that some people post such original thoughts quite frequently. If I thought he'd win, I'd love to see him (Gore) run. However, I don't think he'll win, so I think that if he ran it'd just detract from his environmental message. As for Kerry, I didn't like him in '04 (although I preferred him to Bush), and I don't want to see him run in '08. Not sure who I do want in '08, but that's how it freakin' should be in 2006!
Of course, many do not. Probably more do not. However, Al Gore (the anti-enviros favorite whipping boy) does. That was one of the things I really liked about him in 2000. He understood technology and respected the environment. (Not that I want him to run in '08. I think it would detract from his current campaign.)
That's not much of a prediction (although as someone else pointed out Bigelow might prove you wrong). Currently, NASA's plan is for a lunar base in 2024. Therefore, even an optimist shouldn't expect one before then. A realist might guess 2030, and a pessimist might guess not in this century. Of course, by definition, I'm a realist. :)
I'm really happy that most processors are now 64 bit. 64 bit means it's easier to access more RAM. More RAM means I can simulate more neurons and synapses in less time.
You could have at least used scare quotes around that last "impeached". I'm definitely not in favor of changing the word impeached to mean what the average American thinks it means. As you said, Clinton was impeached. He just wasn't removed from office.
The better I have learned how to write scripts in bash, the easier it has become for me to quickly write a "script" at the command line. I know, you can do this with Python, etc., but for some jobs it's just easier and quicker to use bash. The more tools you have, the quicker you can solve your problems.
Also, I believe that many default distros don't have Python in them. So, if you're on a strange computer you can't count on it being there. Perl or bash, however, you'll almost always find. (This is the same reason I've tried to improve my vi skills even though I'm an emacs fan. I know, I know. Yes, I said emacs.)
Also, it wasn't scientists who were talking about global cooling in the 70's. It was some of the same media types who now think that "both" sides of the global warming "debate" need to be discussed. I.e., journalists who didn't understand science, but want to sell subscriptions. I challenge you to find a single peer-reviewed article supporting global cooling in the 70's. In fact, you'll find that scientists in the 70's were already warning about global warming.
Actually, the volume of water decreases up to about 4 degrees C, IIRC. After that it begins increasing again, like any well-behaved liquid should. Not that this changes your central argument in any way.
I mean:
Land
~~~ = Sinkholes?
Ice
Land ~~~ = Sinkholes? Ice Just having fun...
Especially when what the report has probably lowered is only the worst case scenario. Hopefully, the worst case scenario will always be lessened as one approaches the date in question.
I had assumed that you actually cared about the issue one way or the other. I didn't realize you just had an axe to grind. Carry on.
But please RTFA and not just the summary/headline.
and B) I hope you're familiar with the concept of over-generalization? I.e., you can't extract very much meaningfully predictive information from a single example.
Normally, I'd be happy to argue the actual politics with you (rather than just the logic), but I know it would be fruitless (and that's not intended to be an insult in any way - it's admittedly fruitless in both directions), and I'm a little tired.
Not that I really expect the logic/math argument to be that much more successful... :)
I said we should argue the logic (i.e., the lack of meaning in the denominators) rather than the politics since we definitely won't agree on the politics (at least not completely). If there are X terrorists of type A, and Y terrorists of type B, then the proper sampling would be X/Y regardless of how many total people there are of types A and B - unless you have additional information that we didn't even discuss.
Here's the basic logic: let's assume you have the accurate P(T|A), where T=person is terrorist, and A=person is of type A, and P(T|B). In this case, we'll say (for sake of argument) that P(T|A) = 25/100,000,000 and P(T|B) = 25/1,000,000. I.e., P(T|A) = 0.01 * P(T|B). Now, let's assume that the probability of flying on an airplane is identical for groups A & B (i.e., I'm assuming no "a priori" information). That means that P(F|A) = P(F|B), where F=person is flying on a plane. Let's also assume (although I won't argue it) that P(F|A) is independent of P(T|A). This assumption of equality will mean that we won't actually need to use this probability.
Now, the question you're really trying to ask, when you're (sort of) asking what is the ratio of A to B that should be sampled is what is P(A|T) relative to P(B|T)? Now, Bayes' theorem tells us that P(A|T)P(T) = P(A,T) = P(T|A)P(A). Solving this for P(T) gives us P(T) = P(T|A)P(A)/P(A|T). By similar logic, P(T) = P(T|B)P(B)/P(B|T), so P(T|A)P(A)/P(A|T) = P(T|B)P(B)/P(B|T). Now, if we assume that the ratio of P(A)/P(B) = 100 (i.e., there are 100 times more A than B), and given our initial assumption that P(T|A)/P(T|B) = 0.01, we have that P(A|T) = P(B|T). I.e., for every A you search, you should search 1 (and not 100) B. Note: this means that you will be effectively sampling B at 100 times the rate of A.
Just to argue the logic and not the politics (because I'm certain neither of us will budge on that), your logic is flawed. First of all, that 25 should be 19 (unless I'm missing 6 from somewhere, which I'll admit is possible). So, we'll just let both numbers be 19. (If I really wanted to, I could come up with scores of "Columbinish" events, including one that happened at a school where I used to teach, but I won't go there.) As I'll try to explain, the denominator in the first paragraph is irrelevant for your profile rate in the second paragraph. If there are 100 times more "Whities" than "Muzzies" (gak!), then if you search one W for each one (not 100) M, then you'll be targeting them 100 times as much (since there are 1/100th as many to search). If you searched 100 M's for every 1 W (as you suggest), then you are targeting them 10,000 times as much despite your assertion that they are only 100 times more likely to be a terrorist.
Then, of course, there's the problem of identifying M's. Are we going to use crescent armbands or something? (Does that count towards Godwin?).
Yes, but how many of the Oklahoma City terrorists were Muslims? How many of the abortion clinic bombing terrorists were Muslims? How many of the Columbine terrorists were Muslims?
I can't hear it. If I turn my speakers up, I can hear the distortion in my speakers, but I'm not really hearing it. I know, because when I turn them down, I don't hear anything, but one of my colleagues can, and yells at me to turn it off. :D Hm, let me try that again. Yep, still works. :D (I only play it for a second or two. I'm not that mean.) Actually, from about 20-30 there's a significant drop-off in who can hear it, so don't be surprised if you're not really hearing it. Get a similarly aged (or younger) friend to help you verify whether or not you're hearing it.
And yes, there are people over 60 who can hear it, too. Very few, however.
Actually, I define "passing on the right" the same way you do, but I define "passing someone to your right" to mean that the person you are passing is to your right. So, yes, we're basically agreeing.
There's no doubt that different regions have different types of drivers, and I agree that if you're not passing someone you should get over, but how exactly do you define "substantially faster"? IMO, if you're passing someone to your right (and you're driving at least the speed limit), you need not make any apologies for not going "fast enough" for the people behind you. Sometimes this is complicated somewhat by hills and tractor trailers, of course. (I live near the Blue Ridge Parkway, where this might be a bigger issue than in Massachusetts.)
Also, I should confess that these days I only get on the highway about three to five times a year. I used to commute ~20 miles to work each way when I lived in Atlanta (~5 years ago), but now I commute <1 mile to work each way (via foot) now that I live in glorious Charlottesville.
But seriously, I'm fairly certain they are highly correlated. I drive the speed limit, and I'm always in the correct lane. It is very, very rare that I have had to pass someone going slower than the speed limit, but when I have - they are in the correct lane probably close to 99% of the time.
If you think otherwise, you're probably suffering from confirmation bias.