It only works on a CPU that supports hardware virtualization (Intel's VT or AMD's AMD-V). So the low/mid range processors a business has might not work unless they are quite new. I suppose if they're buying all new Windows 7 boxes they can certainly use XP mode. But if they're upgrading existing machines, it might not work.
Google is awesome for letting people search the web. Facebook is ultimate evil for letting you search Facebook. If you posted information publically, that all your facebook friends can read...in what way did you have an expectation that your facebook friends wouldn't read it? Were you hoping it would get lost in the flood of bullshit and nobody would read it? Really? You were relying on signal:noise ratio for privacy, rather than actually sending a PM? That's beyond absurd.
It doesn't rate different sections differently. But, it works by going, for each sentence, find the most similar sentence and compute the score. The score for matching sentences will be somewhat high. The score for dissimilar ones will be quite negative. So, even if it's 50% recycled stuff, and 50% all new stuff, the negative score from the 50% of the sentences that are new will dominate the total score, and it's going to end up pretty low overall. But yeah, it's not a plagiarism detection algorithm, it's a similarity measure. You can easily trip it up if you're trying to. And you can be measured as "similar" without plagiarism. It's up to a person to make that call. This is just a measure of similarity, originally meant to speed up a literature search. If you took an existing problem, built on previous works, made some improvements and adjustments to previous methodologies, and got better or more interesting results, then your paper probably will have a high similarity score to previous papers. And it's supposed to, that is the entire point of the program! As a side effect, it notices cases of blatant plagarism, where somebody lifts the majority of a paper verbatim and publishes it elsewhere. Though, most cases are "self plagiarism" where a person publishes multiple times with minor variations to inflate their publication list and get more funding.
"Syntactic matching appears to be exactly what this program is doing". At least you openly admit that you are only assuming you know how the fuck it works. Given that they are working in Bioinformatics, and that it's called "ET BLAST" I'm going to go out on a limb and say that it works similar to how BLAST works. When you computing the similarity matrix for a protein (or DNA), well, you could just put those two amino acid sequences (or basepair) side-by-side and count up where they match. Only, some amino acids are quite similar, so replacing one with the other probably wouldn't change the shape of the protein (at least, enough that it doesn't perform its purpose). But some, (like proline and just about any other aa) are quite different, and so aren't too likely to be aligned if they are matched like this. So, you have a pairwise lookup table for likelihood scores. These are usually based on the log-odds ratio. Mean, the log of Pr(X and Y are related) / Pr(X and Y are randomly selected amino acids). The bottom just comes from the common distribution of all 20 aa, and the top comes from wherever you like. Probably also from observation. Now, that's easy! You just sum up the lookup values for all pairs aligned between sequences A and B. But, subsitution is not the only kind of mutation you can get. You can have amino acids cut, and have amino acids added in. So, you assign a cost for these gaps, and score for the aligned parts, plus the cost of the gaps. Only, now that's not trivial. Now it's an NP Hard search problem. BLAST works by an approximation of this. It finds "anchors" which are highly similar subsequences. Then it forces those to be aligned, and has broken things into a several smaller subproblems. Since it's an exponential search problem, even just cutting it in half due to a SINGLE anchor match speeds it up immensely. (Usually you'll only even bother scoring if you have several anchor matches).
Anywho, if you would read not just TFA, but TFP (The fine paper, not the blurb about the paper). (Paper title is "Text similarity: an alternative way to search MEDLINE"), you can see how it works. It does what I guessed before reading it (based only on the name!) it does dynamic programming to align words, and rank them by a log-odds lookup table trained on a large dataset. They do it in a hierarchical manner. They have a whole-text alignment, and a sentence based alignment. That is, in the sentence based one, each sentence is matched up with the highest scoring sentence in the document you're matching it to, and vise versa. And each sentence is scored via the log-odds score. The score of the document is the sum of these scores. And the whole text score aligns everything, so it doesn't rearrange the sentences. In a math-paper context, your common phrase "Let p be an odd prime number" would have a very low score. Because even if all 7 words did line up perfectly, each one is a common word in a math paper, so it would have a low log-odds score. (That is, prime would appear in a lot of sentences that are totally different, so seeing prime is a very poor indicator that two sentences are related at all). So, although it would be a "perfect" match in two different papers, the sentence score would be nearly zero, so while that sentence would not harm the score of the two papers as a whole, it also would NOT be contributing to their "sameness" very much. But yeah, it's still context free, which sucks for natural language, but what are you going to do? You could make it like, a 3rd level context sensitive frequency count, so then your sentence "Let p be an odd prime number" becomes "NULL NULL Let" "NULL Let p" "Let p be" "P be an" "be an odd" "an odd prime" "odd prime number" "prime number NULL" "number NULL NULL". And, you compare the odds of each of THOSE tokens in similar sentences, to their overall occurrence across all sentences. This makes your lookup table a lot bigger, and it means you need a bigger training data set for computing that table in the first place. An
His mistake was calling it "brute force". But, you can find the optimal path in a lot less than the theoretical MAXIMUM required time of n factorial. In 2001 they solved it for 15,000 nodes in 75 days (using 110 500MHz Alpha processors, so 22.6 CPU years). One 150th of the size, with a CPU twice as fast. I'm not so sure it would take a cellphone too long, really. You can find web apps out there that solve 24 nodes in a few minutes, in your browser, USING JAVA SCRIPT.
The thing to remember is that 2^n (not n!) is the worst case upper bound. Meaning, there exist instances that require (at most) that much time to solve. This measure makes no judgment on whether or not such instances are particularly common. In fact. remember how I said somebody solved 15,000 nodes in 22.6 CPU years? (based on a real map). Years later somebody solved it for 88,000 nodes in about 100 CPU years. 5 times the CPU time, 5 times the nodes. Seems more like linear than it does factorial. That's what the OP was talking about, does the graph include sections designed to trip up a greedy algorithm. The reasons the problem in general is so hard is because there CAN be instances that are very hard. They are NOT common. I mean, honestly, if you can do 24 in a few minutes in JavaScript, pulling distances from the Google Maps API, you really thing that most instances are that impossibly hard to do for 100 nodes?
Further, you can approximate to within 1% in a few SECONDS on a Cellphone. (Not all instances, but again, the tricky ones are not common, and possibly never occur in real maps, especially not a bee's flight map where the distances are very close to euclidean)
When asked about the motorcycle's
registration, J.J. told Deputy Duesler that he had purchased it from a guy he knew only as
Skye for $200 three or four days before. J.J. did not have the pink slip. He was also unable
to provide Deputy Duesler with a phone number or address for Skye. An inspection of the
motorcycle showed that it had been modified to operate without a key. Deputy Duesler
arrested J.J. after he determined that the motorcycle was stolen.
How, in the story as he told it, did he meet Skye to buy the motorcycle? He doesn't say. Because he didn't testify. He didn't go to trial, he plead out in exchange for (stricter) probation. All he said on the matter is that it was a person named Skye, and he doesn't know where Skye lives or how to contact him. In case you still doubt, but don't feel like reading the whole decision:
There is nothing in the undisputed record to suggest J.J. used instant messaging or social
networking sites to obtain the stolen motorcycle or drugs. Thus, the probation condition
suffers from the same defect as Holm, Stevens and the ban on J.J. using the Internet for non-
school-related purposes.
How did he arrange to receive the stolen motorcycle
He didn't, not really. He was found riding a dirt bike which had obviously been hotwired. He said "Oh, didn't noticed that, I bought it from some guy". I.E. he stole it, but they can't prove beyond reasonable doubt that there was no "some guy", but at least can prove beyond reasonable doubt that he knew it was stolen (no key, hotwired).
Who said ONLINE play? They are banning people from Single Player. And you are wrong. You bought a license. If you go to the store and buy a box of Office, and you use Word to write a bad review of Windows 7, you have violated the EULA (which says you are not allowed to criticize Microsoft) and therefore your installation of Office is a copyright violation. You see, though the Copyright Act says you are allowed to make as many copies are are necessary for the intended use of the product, the courts have declared that since "everybody knows" you cannot buy software, only license it, the Copyright Act does not apply because you have made no purchase.
Correct. Judges have already thrown out the Copyright Act. Though the Copyright Act explicitly says that it does not apply to any copies necessary for the intended use of the product, the courts have declared that this clause is unfair to copyright giants, and ignore it.
It's not the RAM part. Activision (there is no such company as Blizzard anymore) is saying it's illegal to play a game, because playing a game loads it from your HD and that is a copyright violation except where you are explicitly granted permission to "distribute" your "copy" into RAM by Activision. Therefore, if you cheat at SINGLE PLAYER they have the legal right to remotely delete your copy. To ban you from SINGLE FUCKING PLAYER. And people said it would be just fine when Blizzard killed offline play utterly by removing LAN and single player from the game.
The most damning evidence against the "doomsday" is the fact that at least one Mayan king wrote about how he hoped people will still celebrate him in, then he gave a date, a date several thousand years after the end of the Mayan long calendar. So, did he not get the memo that the world would have been destroyed thousands of years before that date? Was he just oblivious? Seems contrary to assume he was clueless, since the only reason we think the Mayan calendar predicts the end of the world is the assumption that they were all-knowing and all-seeing, by virtue of not being us.
I've never heard of anything other than 16 bit software that wouldn't run under Win7 64, but runs fine under Win7 32.
Re:The industry can take all the time it needs
on
WD Launches 3 Terabyte HD
·
· Score: 2, Informative
I was working at a government research center in the 90s. Their main server had a 4 drive raid 1 array. One day there was an emergency shutdown because 3 drives suffered near-simultaneous head crashes. I think nobody told the IT guy you are supposed to make sure all your RAID drives came from different batches. Either that or somebody kicked the damn thing and didn't own up to it;) They did regular tape backups, too, so it wouldn't be the end of the world. Just maybe a day or two of work lost. Well, if you only use your network drive, rather than using your network drive as a backup;)
Section 64 of CEPA 1999 defines a substance as toxic "if it is entering or may enter the environment in a quantity or concentration or under conditions that: have or may have an immediate or long-term harmful effect on the environment or its biological diversity; constitute or may constitute a danger to the environment on which life depends; or constitute or may constitute a danger in Canada to human life or health." That's the definition they use. It is up to the government to determine what is a "long-term harmful effect" on human life or health. They have conducted studies indicating it is acutely toxic (determined by LD50, like you suggest) to aquatic life, so that's "immediate or long-term harmful effect on the environment or its biological diversity". Also, it's been found, by Canadian, US, and EU studies, to have a significant effect on fetal and childhood development. Specifically on the development of the brain and prostate gland. This makes it a "reproductive toxin" as described in CPR (Controlled Product Regulations) sections 55 and 58. This is not done by LD50, but by evaluation of studies as to whether or not current studies indicates "Evidence of a physiological effect". They looked at those Canadian, American, and European studies, and concluded that there is in fact evidence of physiological effect. So, while it is not toxic in the sense of having an LD50, it is toxic by their definition of having a physiological effect on fetal and child development. Anywho, here is the Environmental Protection Act that defines that stuff. And here is the actual announcement from the government, which you can reach from TFA if you follow enough links.
The real question is, when they say "See things your friends like" do they mean, Bing will be inserting "Like" buttons into all of their shopping pages and links, so you can "like" a product in the Facebook sense? Or do they mean it in the Facebook Beacon Sense: "Bing will now report all purchases to Facebook if you forgot to sign out, and when you visit Bing you will see a list of products your friends have bought, without their knowledge since this is opt-out, in the sense that you cannot opt out, but after the broadcast to your friends has been made, you have the option to delete it again!"
Like I said, you are probably thinking that Monty Hall is picking at random, so it's only by chance that he didn't open the door you already picked. That's wrong. The rule is, to heighten the excitement, he opens one of the "bad" choices you didn't make. He can never open the door you picked. If he could, then you are right, it's 50/50. But, he can't open the door you picked, and he can't open the door the prize is actually behind, either.
You are saying two things. You are saying that the chance you picked the right door is both 1/100 and 1/2. How can it be both? Go back to 3 doors since obviously going to 100 didn't help you to think clearly on the problem. There are 3 cases, A B C. A=car behind door 1, B=car behind door 2, C = car behind door 3. Let's always pick door number two, for starters.
In A, we guessed wrong. Car behind 1, we guessed 2, so Monty Hall opens 3 and shows a goat. Staying with door two gives us a goat.
In B, we guessed right. Car behind 2, we guessed 2. Monty hall can open either 1 or 3. But either way, staying with door two gives us a car
In C, we guessed wrong. Car behind 3, we guessed 2. Monty Hall opens door number 1 and shows a goat. Staying with door two gives us a goat
As you can imagine, the play-by-play is symmetric so this represents all cases if you just change the labels on the doors in your head first. So, if you assume that the situation is equally likely to be A, B, or C, then you can see that in only one case does staying give you the car. 1/3. Because deciding to stay is deciding to assume that your initial guess was correct. Your initial guess is only correct with probability 1/3. Or, 1/100 in the 100 door case. You cannot get the car by staying unless your initial guess was right, so the odds of getting the car by staying are the same as the odds of guessing correctly initially.
However, this is only true of Monty Hall always gives you the option. If he has a choice of whether or not he gives you the option of switching, then you can make the argument that switching is the worse choice. After all, if you guessed right, switching means you lose, and if you guessed wrong, switching means you win. Since odds are against you picking right initially, Monty Hall's best bet would be to give you the option only if giving you the option makes you lose. In that case, you've picked right and have won. Monty Hall always loses if he doesn't give you the choice, so giving you the choice then doesn't hurt him. And conversely, he should never give you the option if you're picked wrong. Because then he's taking you from "100% lost" to "might win" so that's a poor move on his part. It's never 50/50, though. If you always get the offer, it's best to switch. But if it's Monty's choice, its best to never switch, because that will basically mean you always lose. (He might occasionally make the offer when you've picked wrong, just to maintain the illusion that it's random and switching CAN be a good choice, so it's not 100% a loss to switch, but it's way worse than 1/3). So, if Monty Hall always offers, or if he picks randomly without knowing himself which door has the car, you are better off switching. But if he decides whether or not to give you the option of switching, then you are better off sticking with it, because he'd only give you the option to get you to switch away from the right door!
Your cat and mouse example is a perfect example of exactly why 0.9999... = 1! Cat and mouse are 1 m apart, cover half the distance, 1/2m, 1/4m, 1/8m...infinity later, and sum them all up, and this is equal to exactly 1m. The reason the cat cannot reach it is because it DOES require infinite steps, and the cat cannot take infinite steps. However, 0.9999999.... has infinite digits already. You cannot have a handful of 0.9999.... cups of something, because it requires infinite steps to produce. But on paper, you can have such a number.
Put another way, if 1 is greater than 0.9bar, then what is 1 - 0.9bar? 1 - 0.9 = 0.1, or, ten to the minus one. 1-0.99 = 0.01, ten to the minus two. AKA, one minus (zero followed by x nines) = ten to the minus x. How many nines, again? Infinity. OK, there is your answer. The difference here is ten to the minus infinity. How do we handle that power? Who cares. What does it mean? It means infinite zeros followed by a one. In your example, the cat cannot reach the mouse because, if the space is divided into an infinite number of "allowed" steps, it cannot cover all infinite steps in its finite lifetime. By the same argument, if there are an infinite number of zeros, there is no one at the end. You with your infinitely accurate measuring device would spend your entire life SURE that if you just keep going, it will eventually say "1", but it won't, not ever. 0.0000000000000000000000000000.... is just 0. 1 - 0.9bar = 0, so they are equal.
How does 10 x 0.99999999.... = 9.9? How did you erase, as you say, infinity significant digits? Nice parlor trick. So since you're so sure that math is bogus and 1 != 0.999..., then what is 1 - 0.999. If they are not equal, it must be a value greater than zero. What is that value? Zero point zero zero zero....infinity zeros later, 1? How can there be a digit after INFINITE zeros? There is no end to the zeros. It is you, good sir, that does not truly understand that it has infinite digits.
The DEA and the FBI aren't the same agency. Because the DEA decided to play fast and loose with a GPS tracker doesn't imply that the FBI would do the same thing. Their policy is to get a warrant for a GPS tracker, and I don't think they would have amended it to "except in these states..." that quickly. Whether agents followed proper procedure or not is another matter. But again, the point is there is no indication that they didn't have a warrant. Nobody in the media seems to be interested in even asking, they would rather make not-exactly-false statements like "in the wake of the ruling..." and "they did not require a warrant thanks to the liberals!".
"No warrant required," doesn't necessarily mean they didn't have one. The truth is, other than showing up to pick it up, the FBI has refused to confirm or deny that he was ever under surveillance. You can suspect that after the ruling saying they don't need warrants anymore, that they didn't have one. But, they asked him questions and he lawyered up 6 months ago. The device could have been planted before the ruling was ever made! Given that, I don't think it's a particularly safe assumption that they didn't get a warrant, in the hopes that an appeals court would overturn the requirement at a later date.
Slashdot Mirror
It only works on a CPU that supports hardware virtualization (Intel's VT or AMD's AMD-V). So the low/mid range processors a business has might not work unless they are quite new. I suppose if they're buying all new Windows 7 boxes they can certainly use XP mode. But if they're upgrading existing machines, it might not work.
Google is awesome for letting people search the web. Facebook is ultimate evil for letting you search Facebook. If you posted information publically, that all your facebook friends can read...in what way did you have an expectation that your facebook friends wouldn't read it? Were you hoping it would get lost in the flood of bullshit and nobody would read it? Really? You were relying on signal:noise ratio for privacy, rather than actually sending a PM? That's beyond absurd.
It doesn't rate different sections differently. But, it works by going, for each sentence, find the most similar sentence and compute the score. The score for matching sentences will be somewhat high. The score for dissimilar ones will be quite negative. So, even if it's 50% recycled stuff, and 50% all new stuff, the negative score from the 50% of the sentences that are new will dominate the total score, and it's going to end up pretty low overall. But yeah, it's not a plagiarism detection algorithm, it's a similarity measure. You can easily trip it up if you're trying to. And you can be measured as "similar" without plagiarism. It's up to a person to make that call. This is just a measure of similarity, originally meant to speed up a literature search. If you took an existing problem, built on previous works, made some improvements and adjustments to previous methodologies, and got better or more interesting results, then your paper probably will have a high similarity score to previous papers. And it's supposed to, that is the entire point of the program! As a side effect, it notices cases of blatant plagarism, where somebody lifts the majority of a paper verbatim and publishes it elsewhere. Though, most cases are "self plagiarism" where a person publishes multiple times with minor variations to inflate their publication list and get more funding.
"Syntactic matching appears to be exactly what this program is doing". At least you openly admit that you are only assuming you know how the fuck it works. Given that they are working in Bioinformatics, and that it's called "ET BLAST" I'm going to go out on a limb and say that it works similar to how BLAST works. When you computing the similarity matrix for a protein (or DNA), well, you could just put those two amino acid sequences (or basepair) side-by-side and count up where they match. Only, some amino acids are quite similar, so replacing one with the other probably wouldn't change the shape of the protein (at least, enough that it doesn't perform its purpose). But some, (like proline and just about any other aa) are quite different, and so aren't too likely to be aligned if they are matched like this. So, you have a pairwise lookup table for likelihood scores. These are usually based on the log-odds ratio. Mean, the log of Pr(X and Y are related) / Pr(X and Y are randomly selected amino acids). The bottom just comes from the common distribution of all 20 aa, and the top comes from wherever you like. Probably also from observation. Now, that's easy! You just sum up the lookup values for all pairs aligned between sequences A and B. But, subsitution is not the only kind of mutation you can get. You can have amino acids cut, and have amino acids added in. So, you assign a cost for these gaps, and score for the aligned parts, plus the cost of the gaps. Only, now that's not trivial. Now it's an NP Hard search problem. BLAST works by an approximation of this. It finds "anchors" which are highly similar subsequences. Then it forces those to be aligned, and has broken things into a several smaller subproblems. Since it's an exponential search problem, even just cutting it in half due to a SINGLE anchor match speeds it up immensely. (Usually you'll only even bother scoring if you have several anchor matches).
Anywho, if you would read not just TFA, but TFP (The fine paper, not the blurb about the paper). (Paper title is "Text similarity: an alternative way to search MEDLINE"), you can see how it works. It does what I guessed before reading it (based only on the name!) it does dynamic programming to align words, and rank them by a log-odds lookup table trained on a large dataset. They do it in a hierarchical manner. They have a whole-text alignment, and a sentence based alignment. That is, in the sentence based one, each sentence is matched up with the highest scoring sentence in the document you're matching it to, and vise versa. And each sentence is scored via the log-odds score. The score of the document is the sum of these scores. And the whole text score aligns everything, so it doesn't rearrange the sentences. In a math-paper context, your common phrase "Let p be an odd prime number" would have a very low score. Because even if all 7 words did line up perfectly, each one is a common word in a math paper, so it would have a low log-odds score. (That is, prime would appear in a lot of sentences that are totally different, so seeing prime is a very poor indicator that two sentences are related at all). So, although it would be a "perfect" match in two different papers, the sentence score would be nearly zero, so while that sentence would not harm the score of the two papers as a whole, it also would NOT be contributing to their "sameness" very much. But yeah, it's still context free, which sucks for natural language, but what are you going to do? You could make it like, a 3rd level context sensitive frequency count, so then your sentence "Let p be an odd prime number" becomes "NULL NULL Let" "NULL Let p" "Let p be" "P be an" "be an odd" "an odd prime" "odd prime number" "prime number NULL" "number NULL NULL". And, you compare the odds of each of THOSE tokens in similar sentences, to their overall occurrence across all sentences. This makes your lookup table a lot bigger, and it means you need a bigger training data set for computing that table in the first place. An
His mistake was calling it "brute force". But, you can find the optimal path in a lot less than the theoretical MAXIMUM required time of n factorial. In 2001 they solved it for 15,000 nodes in 75 days (using 110 500MHz Alpha processors, so 22.6 CPU years). One 150th of the size, with a CPU twice as fast. I'm not so sure it would take a cellphone too long, really. You can find web apps out there that solve 24 nodes in a few minutes, in your browser, USING JAVA SCRIPT.
The thing to remember is that 2^n (not n!) is the worst case upper bound. Meaning, there exist instances that require (at most) that much time to solve. This measure makes no judgment on whether or not such instances are particularly common. In fact. remember how I said somebody solved 15,000 nodes in 22.6 CPU years? (based on a real map). Years later somebody solved it for 88,000 nodes in about 100 CPU years. 5 times the CPU time, 5 times the nodes. Seems more like linear than it does factorial. That's what the OP was talking about, does the graph include sections designed to trip up a greedy algorithm. The reasons the problem in general is so hard is because there CAN be instances that are very hard. They are NOT common. I mean, honestly, if you can do 24 in a few minutes in JavaScript, pulling distances from the Google Maps API, you really thing that most instances are that impossibly hard to do for 100 nodes?
Further, you can approximate to within 1% in a few SECONDS on a Cellphone. (Not all instances, but again, the tricky ones are not common, and possibly never occur in real maps, especially not a bee's flight map where the distances are very close to euclidean)
How, in the story as he told it, did he meet Skye to buy the motorcycle? He doesn't say. Because he didn't testify. He didn't go to trial, he plead out in exchange for (stricter) probation. All he said on the matter is that it was a person named Skye, and he doesn't know where Skye lives or how to contact him. In case you still doubt, but don't feel like reading the whole decision:
He didn't, not really. He was found riding a dirt bike which had obviously been hotwired. He said "Oh, didn't noticed that, I bought it from some guy". I.E. he stole it, but they can't prove beyond reasonable doubt that there was no "some guy", but at least can prove beyond reasonable doubt that he knew it was stolen (no key, hotwired).
It's a book review, not a product review. If you don't know what the subject of the book is, why do you care how good the book is?
Who said ONLINE play? They are banning people from Single Player. And you are wrong. You bought a license. If you go to the store and buy a box of Office, and you use Word to write a bad review of Windows 7, you have violated the EULA (which says you are not allowed to criticize Microsoft) and therefore your installation of Office is a copyright violation. You see, though the Copyright Act says you are allowed to make as many copies are are necessary for the intended use of the product, the courts have declared that since "everybody knows" you cannot buy software, only license it, the Copyright Act does not apply because you have made no purchase.
Correct. Judges have already thrown out the Copyright Act. Though the Copyright Act explicitly says that it does not apply to any copies necessary for the intended use of the product, the courts have declared that this clause is unfair to copyright giants, and ignore it.
It's not the RAM part. Activision (there is no such company as Blizzard anymore) is saying it's illegal to play a game, because playing a game loads it from your HD and that is a copyright violation except where you are explicitly granted permission to "distribute" your "copy" into RAM by Activision. Therefore, if you cheat at SINGLE PLAYER they have the legal right to remotely delete your copy. To ban you from SINGLE FUCKING PLAYER. And people said it would be just fine when Blizzard killed offline play utterly by removing LAN and single player from the game.
The most damning evidence against the "doomsday" is the fact that at least one Mayan king wrote about how he hoped people will still celebrate him in, then he gave a date, a date several thousand years after the end of the Mayan long calendar. So, did he not get the memo that the world would have been destroyed thousands of years before that date? Was he just oblivious? Seems contrary to assume he was clueless, since the only reason we think the Mayan calendar predicts the end of the world is the assumption that they were all-knowing and all-seeing, by virtue of not being us.
I've never heard of anything other than 16 bit software that wouldn't run under Win7 64, but runs fine under Win7 32.
I was working at a government research center in the 90s. Their main server had a 4 drive raid 1 array. One day there was an emergency shutdown because 3 drives suffered near-simultaneous head crashes. I think nobody told the IT guy you are supposed to make sure all your RAID drives came from different batches. Either that or somebody kicked the damn thing and didn't own up to it ;) They did regular tape backups, too, so it wouldn't be the end of the world. Just maybe a day or two of work lost. Well, if you only use your network drive, rather than using your network drive as a backup ;)
Section 64 of CEPA 1999 defines a substance as toxic "if it is entering or may enter the environment in a quantity or concentration or under conditions that: have or may have an immediate or long-term harmful effect on the environment or its biological diversity; constitute or may constitute a danger to the environment on which life depends; or constitute or may constitute a danger in Canada to human life or health." That's the definition they use. It is up to the government to determine what is a "long-term harmful effect" on human life or health. They have conducted studies indicating it is acutely toxic (determined by LD50, like you suggest) to aquatic life, so that's "immediate or long-term harmful effect on the environment or its biological diversity". Also, it's been found, by Canadian, US, and EU studies, to have a significant effect on fetal and childhood development. Specifically on the development of the brain and prostate gland. This makes it a "reproductive toxin" as described in CPR (Controlled Product Regulations) sections 55 and 58. This is not done by LD50, but by evaluation of studies as to whether or not current studies indicates "Evidence of a physiological effect". They looked at those Canadian, American, and European studies, and concluded that there is in fact evidence of physiological effect. So, while it is not toxic in the sense of having an LD50, it is toxic by their definition of having a physiological effect on fetal and child development. Anywho, here is the Environmental Protection Act that defines that stuff. And here is the actual announcement from the government, which you can reach from TFA if you follow enough links.
And remember that most cans are lined with BPA containing plastic.
The real question is, when they say "See things your friends like" do they mean, Bing will be inserting "Like" buttons into all of their shopping pages and links, so you can "like" a product in the Facebook sense? Or do they mean it in the Facebook Beacon Sense: "Bing will now report all purchases to Facebook if you forgot to sign out, and when you visit Bing you will see a list of products your friends have bought, without their knowledge since this is opt-out, in the sense that you cannot opt out, but after the broadcast to your friends has been made, you have the option to delete it again!"
Like I said, you are probably thinking that Monty Hall is picking at random, so it's only by chance that he didn't open the door you already picked. That's wrong. The rule is, to heighten the excitement, he opens one of the "bad" choices you didn't make. He can never open the door you picked. If he could, then you are right, it's 50/50. But, he can't open the door you picked, and he can't open the door the prize is actually behind, either.
So you are saying there is a school of thought that says there is no such thing as one third?
You are saying two things. You are saying that the chance you picked the right door is both 1/100 and 1/2. How can it be both? Go back to 3 doors since obviously going to 100 didn't help you to think clearly on the problem. There are 3 cases, A B C. A=car behind door 1, B=car behind door 2, C = car behind door 3. Let's always pick door number two, for starters.
In A, we guessed wrong. Car behind 1, we guessed 2, so Monty Hall opens 3 and shows a goat. Staying with door two gives us a goat.
In B, we guessed right. Car behind 2, we guessed 2. Monty hall can open either 1 or 3. But either way, staying with door two gives us a car
In C, we guessed wrong. Car behind 3, we guessed 2. Monty Hall opens door number 1 and shows a goat. Staying with door two gives us a goat
As you can imagine, the play-by-play is symmetric so this represents all cases if you just change the labels on the doors in your head first. So, if you assume that the situation is equally likely to be A, B, or C, then you can see that in only one case does staying give you the car. 1/3. Because deciding to stay is deciding to assume that your initial guess was correct. Your initial guess is only correct with probability 1/3. Or, 1/100 in the 100 door case. You cannot get the car by staying unless your initial guess was right, so the odds of getting the car by staying are the same as the odds of guessing correctly initially.
However, this is only true of Monty Hall always gives you the option. If he has a choice of whether or not he gives you the option of switching, then you can make the argument that switching is the worse choice. After all, if you guessed right, switching means you lose, and if you guessed wrong, switching means you win. Since odds are against you picking right initially, Monty Hall's best bet would be to give you the option only if giving you the option makes you lose. In that case, you've picked right and have won. Monty Hall always loses if he doesn't give you the choice, so giving you the choice then doesn't hurt him. And conversely, he should never give you the option if you're picked wrong. Because then he's taking you from "100% lost" to "might win" so that's a poor move on his part. It's never 50/50, though. If you always get the offer, it's best to switch. But if it's Monty's choice, its best to never switch, because that will basically mean you always lose. (He might occasionally make the offer when you've picked wrong, just to maintain the illusion that it's random and switching CAN be a good choice, so it's not 100% a loss to switch, but it's way worse than 1/3). So, if Monty Hall always offers, or if he picks randomly without knowing himself which door has the car, you are better off switching. But if he decides whether or not to give you the option of switching, then you are better off sticking with it, because he'd only give you the option to get you to switch away from the right door!
So what's the answer. Nice total avoidance of the question. What is 1 - 0.999...? You are saying it is a non-zero number. So what is it?
Your cat and mouse example is a perfect example of exactly why 0.9999... = 1! Cat and mouse are 1 m apart, cover half the distance, 1/2m, 1/4m, 1/8m...infinity later, and sum them all up, and this is equal to exactly 1m. The reason the cat cannot reach it is because it DOES require infinite steps, and the cat cannot take infinite steps. However, 0.9999999.... has infinite digits already. You cannot have a handful of 0.9999.... cups of something, because it requires infinite steps to produce. But on paper, you can have such a number.
Put another way, if 1 is greater than 0.9bar, then what is 1 - 0.9bar? 1 - 0.9 = 0.1, or, ten to the minus one. 1-0.99 = 0.01, ten to the minus two. AKA, one minus (zero followed by x nines) = ten to the minus x. How many nines, again? Infinity. OK, there is your answer. The difference here is ten to the minus infinity. How do we handle that power? Who cares. What does it mean? It means infinite zeros followed by a one. In your example, the cat cannot reach the mouse because, if the space is divided into an infinite number of "allowed" steps, it cannot cover all infinite steps in its finite lifetime. By the same argument, if there are an infinite number of zeros, there is no one at the end. You with your infinitely accurate measuring device would spend your entire life SURE that if you just keep going, it will eventually say "1", but it won't, not ever. 0.0000000000000000000000000000.... is just 0. 1 - 0.9bar = 0, so they are equal.
How does 10 x 0.99999999.... = 9.9? How did you erase, as you say, infinity significant digits? Nice parlor trick. So since you're so sure that math is bogus and 1 != 0.999..., then what is 1 - 0.999. If they are not equal, it must be a value greater than zero. What is that value? Zero point zero zero zero....infinity zeros later, 1? How can there be a digit after INFINITE zeros? There is no end to the zeros. It is you, good sir, that does not truly understand that it has infinite digits.
The DEA and the FBI aren't the same agency. Because the DEA decided to play fast and loose with a GPS tracker doesn't imply that the FBI would do the same thing. Their policy is to get a warrant for a GPS tracker, and I don't think they would have amended it to "except in these states..." that quickly. Whether agents followed proper procedure or not is another matter. But again, the point is there is no indication that they didn't have a warrant. Nobody in the media seems to be interested in even asking, they would rather make not-exactly-false statements like "in the wake of the ruling..." and "they did not require a warrant thanks to the liberals!".
"No warrant required," doesn't necessarily mean they didn't have one. The truth is, other than showing up to pick it up, the FBI has refused to confirm or deny that he was ever under surveillance. You can suspect that after the ruling saying they don't need warrants anymore, that they didn't have one. But, they asked him questions and he lawyered up 6 months ago. The device could have been planted before the ruling was ever made! Given that, I don't think it's a particularly safe assumption that they didn't get a warrant, in the hopes that an appeals court would overturn the requirement at a later date.