I didn't realize that state compacts had to be approved by Congress. I suppose it makes sense under the Interstate Commerce clause, but I've been unable to find any good cites on this. Can you point me in the right direction?
I'd be curious to see how enforceable the contract turns out to be. I can imagine a state changing its mind midway through the voting, or secretly changing its vote, or something. If the other states sue to enforce the contract, would it prove valid?
It does make recounts rather a mess. One advantage to the electoral college system is that as messy as the Florida recount was, at least it was in only one state. The election of 2000 was very close even in popular terms, and without the electoral college every single state would have ended up having a recount, because every single vote would matter. But gosh, other countries manage to work it out.
The states that have already talked about signing on are big states: California, New York, Colorado, Illinois and Missouri. States who are under-represented in the electoral college. The little states, who currently benefit from having their individual votes be worth nearly 3 times as much as a voter from California or New York, will pitch a major hissy fit.
I haven't run the numbers, but I suspect that such a scheme will tend to favor Democrats over Republicans, at least with the current distributions. Those small states tend to be red states. Certainly the one recent example where one can point to a candidate getting an advantage from the electoral college favored a Republican over a Democrat, so any attempt to swing it towards a proportional vote will be greeted in red states as an attempt to make it more blue.
Hey, sorry if I offended. I was trying to make a joke; I apologize if it landed with a thud. I have several friends who are LD in a variety of ways, so I sympathize.
(What I could use is a "bad taste" filter on my posts. This is the fourth time in the last month I've stuck my foot in my mouth in public.)
1. The Box Model, is the math 9in? IE finally not backwards from every one else, does it now make sense? Will 'Border' not be full scree when I just set them to '30px'?
2. Oh one thing I am happy about in Fire Fox that is a long time coming for me is the spell check,
We noticed.
Actually, "scree" is a perfectly valid word, and OpenOffice doesn't flag "9in" (or "30px" for that matter). And it's sure not going to tell you that "Firefox" is one word.
But really, you're right: a spell check will probably prevent a lot of typos from getting out into the world. I've had to fix a few of my own by hand in typing this, and nothing kills a "spelling nazi" joke like screwing it up yourself.
I guess what they're getting at is that IE has a button to let you bring up the bookmarks toolbar, rather than having to go to the menu or use a keyboard shortcut.
Or maybe they're referring to the "Add..." button on the Favorites sidebar.
Either way, I'm sure that if anybody cared enough to create this feature it would be an extension. It sounds like the sort of thing you do as part of a tutorial titled "My First Incredibly Easy Extension". It hardly sounds like a feature on par with a pop-up blocker. And in fact given that it's a button I'd never use I'm just as happy that it's not sucking up screen real estate on the navigation toolbar.
It just goes to show you that one person's completely useless screen-clogging waste is the next person's critical can't-live-without feature.
On the font-weight vs. text-decoration distinction, CSS is inheriting the terminology from typesetters and type designers. Technically, "Times Roman" and "Helvetica" designate "typefaces". "Times Roman Bold" and "Helvetica Italic" are "fonts". It's a property of the design itself; the bold and italic aren't simply automatically-derived versions of the typeface but require an artist to sit down and design them separately. (Some even incorporate the size; Times Roman 24 isn't always just a zoomed-out version of Times Roman 12).
(It gets even more complicated with the notion of "font families", but I don't understand the distinction there, either.)
Underlining, on the other hand, is just something you do to it; there isn't any "Times Roman Underlined". That makes it a property of the text, not of the font or face. You don't need a designer to add it.
It sucks that you need such details to do something that you get just by pushing a button in every WYSIWIG word processor in the world. What we need, and what I haven't seen yet, is a WYSIWIG designer for CSS. I envision something equivalent to what Word and OpenOffice call "character styles", but frankly most people don't use them even when they're available.
And Word/OpenOffice still lack (for the most part) an equivalent of CSS layout, which is the part I still find hard. As you point out, CSS's box model seems to be missing some really basic ideas, and that causes many people to just say, "This is 300 pixels wide and it looks fine at a font size I'm comfortable reading and I don't want to f*** with it any more."
In theory, that's why the cable/telcos want to break "net neutrality": they want to charge Apple to ensure the bandwidth.
What drives the people campaigning for net neutrality is the potential for abuse: cutting or drastically throttling traffic from any site who won't pay you, no matter how little traffic they actually generate, or perhaps sites whose politics the ISP disagrees with.
People are buying the TV shows at 320x240 and watching them on their iPods. It stuns me, but the convenience seems to outweigh the resolution issues for a lot of people. For movies you've got the additional hassle of the different screen shape; perhaps the latest round of iPods have a 16x9 screen.
We see exactly the same argument whenever there's a HD article: lots of Slashdotters claiming that 640x480 is plenty for them and they don't feel the need to spend a lot of money on a new TV and player for the higher resolution.
The article says that they had "not paid federal wagering taxes on $3.3 billion on wagers taken by the firm". At a guess, if the house is taking a 2% cut, at a 30% tax rate, that's almost $20 million that they owe. That's enough money to justify arresting somebody and having them extradited.
(There's a whole big deal going on between the US and Britain with respect to extradition, having to do with levels of proof, but I'll assume for the moment that they've got some proof on this guy. It's a separate argument.)
I've never been exactly sure what "racketeering" is. It's kind of the US-white collar equivalent of what the Soviets used to call "hooliganism": a generic crime applied to people doing scuzzy stuff in bulk where you don't want to (or can't) pin any individual crimes on them.
The charge makes me nervous, because its vagueness makes it subject to abuse. But tax evasion is a crime I an understand. There's also the fraud charges, which is a more specific crime but I don't see specific allegations in the article. Presumably they're in the indictment. It lets them avoid having to test their dubious bans against internet gambling directly.
It's not really a question of "enough". It's a question of what a stock is worth.
A naive interpretation of a stock's worth is that you take the company's assets, sell them, and divide them up among the shareholders. But a company is clearly worth more than that: it makes profits. So you should add to the worth the amount of money you could expect to make from it if you were to divide up the profits.
But what are the profits going to be? That's hard to say, and much magic goes into figuring it out. People get some idea of what they think it's worth, and they price the shares accordingly. Their guesses aren't secrets: the company (sometimes) says what they think the profits will be, and so do analysts, and they share them.
Make less than that, and the share price goes down. It has to: that share is worth less than you thought it was. It doesn't matter if it's still profit; your share of that profit is less than you thought it was going to be, and you'll pay less for it. And if you extrapolate form how much they missed this profit goal to the future ones that you already had figured in when you made your stock price, you'll pay even less.
This is a vast oversimplification, sadly. The price goes down by more than the profit miss indicates because there's an overcorrection: people see others selling the stock and so they figure it's going down and sell theirs, too. Predicting that herd behavior is obnoxious, but if you were to buy-and-hold the stock for a long time the overcorrections will even out over time. Or you can day trade and take advantage of those overcorrections on a daily basis, but if you do your first investment had better be in Tums. In bulk.
In other words: the share price includes a "discount" of what they expect future earnings to be. Miss those earnings, and you have to recompute the discount. An actual decline would be a disaster, but even less-increase affects your calculations.
I believe you're correct. I'm under the impression (and I haven't read the decision) that the judge wasn't exactly encouraging kinderstart to re-file. He was just pointing out to Google that there was still an avenue for this to keep coming back at them, even while he was rejecting kinderstart's original claim of antitrust.
That's correct. Kinderstart would further have to show that a pagerank is some kind of objective fact, not just an opinion. I detailed this in a response to a sister posting to yours, but the gist is that Kinderstart could conceivably demonstrate that a low rank isn't merely "you suck" but "you are worthless", and they'd refute that statement by showing that there is useful information on their site. Google gave them a 0, and they'd argue that connotes "completely worthless", which is closer to an objective statement of fact than merely "you suck".
I doubt it would work, and I suspect the judge does, too, but he didn't want to rule on that argument if it hasn't been made yet. As long as it has a chance of working, he has to permit it to happen.
They'd have to show that Google is more than just saying "You suck". It's perfectly legal to express an opinion. Even if they can prove malice, their next step is to prove that the rank is a false statement of fact.
So the judge is saying that he's also not completely closed to that idea, either. Google would probably say that a page rank is a matter of opinion. They'll respond that it's not; it's an objective number, and you artifically lowered ours. Google will say that the number is a product of an algorithm, and that the algorithm encodes certain opinions that they have, including "You suck."
Or kinderstart could say, "Google makes a statement that our site is less 'good' than site X. We submit a bunch of paperwork to show our site is better. Therefore their statement is false, and we show malice with the artificial page-rank lowering."
I doubt the judge would buy that, either, if for no other reason than that it makes every single Google decision actionable. But he didn't want to rule it out completely with prejudice.
An RFID chip is really, really, really simple. So simple it doesn't have any power on board; it just uses the power of the radio you beam at it. And so simple it costs less than a penny to make. You're never going to get a fingerprint reader on one.
The judge says he's open to the idea that if Google manually changed the pagerank to zero, then maybe it's defamation, which is a valid reason to bring a suit. It's equivalent to saying, out loud and deliberately, "You suck". He's saying that if they do it manually, that could be evidence for malice, which is a requirement for a defamation suit. (At least according to Wikipedia; IANAL.)
I suspect that the judge wanted to leave the idea open, since it hadn't been explored completely and therefore he couldn't absolutely rule it out, but I doubt it would fly. First they have to show that malice, and I can't help but think that they'll have a hard time with that. They'd need somebody with inside knowledge of the decision process; the judge has pretty much said that the lower number is not in and of itself evidence of malice.
I'm sure they've got something they'll throw at this, so I doubt it's the last time we've heard of them. I suspect from here it'll be:
Even the notion that "constants are constant" is just a special case of the assumption that there are fundamental laws, and that they tend to be relatively simple. The constants appear, more or less by magic, in formualas. That magic is a source of wonder, and often the whole formula is replaced by a radically different one, from which the constant falls out as a good approximation from the limited range of experiments we've been able to do so far.
But the new formula usually has constants of its own. There's a non-rational belief that sooner or later we'll reduce it all to one very small set of constants that seem less like magic, but I can't prove it.
To my mind, the difference is in the nature of the principles. A scientist's principles are very short and simple, and pretty much everthing derives from there. They're so simple that they seem pretty much obvious. (Though in general "obvious" is not a good thing; many obvious things are wrong.)
By contrast, religions usually have vastly complicated sets of principles when it comes to the nature of the universe, usually a God with a mind whose decisions can seem arbitrary and capricious, and that same God also imposes a rather complex set of moral principles which cannot be derived from the physical properties of the universe. They often further posit punishments and rewards for those.
Religions believe that the world is a very complex place. Scientists believe that the world is, fundamentally, a place with simple rules that appear complex through interaction. That simplicity is very compelling to me, but I can't really prove it.
Yeah, I noticed the same thing. In one sense it's kind of irritating to have the insinuation perpetrate the myth that scientists have a non-rational belief equivalent to a religious belief, and that these scientsts are some kind of heretics. We know what they meant, but still...
A more precise headline is somewhat harder to write: "Scientists find evidence that they may have to refine or even refactor some really, really well-demonstrated theories" isn't nearly as punchy.
(Scientists do, in fact, have non-rational fundamentally held beliefs, but they're nothing so simple as "Einstein was right, Darwin was right". Trying to convince somebody that a scientist's real religious belief is "The universe has some sort of fundamental, objective, and probably comparatively simple law, one that we can understand or at least produce successively more acurate approximations, one that can be modeled mathematically and is true over all space and time, one that makes predictions that can be tested and will stand up to all such tests all the time" is rather more complicated and less fun. And yes, I recognize that my approximation of that belief above is both more complicated and less accurate than some other formulations, but I'm already drifting dangerously off-topic.)
Ultimately, very little. At core, they're probably identical techniques, and if I were reviewing this as a scientific paper I'd ding them for not answering exactly that question. There are such strong parallels between the two (train them on known data, add up probabilities, cut stuff on a threshold) that I strongly suspect that they're identical.
There are useful things to be gained from a change of metaphor. For example, one difference between this and most bayesian spam filter implementations is that this explicitly incorporates a decay function. That could be useful, if a word that used to be common in spam no longer is (e.g. if I actually decided to buy a Rolex, it's no longer a strong spam indicator, whereas right now any email mentionining "Rolex" is 99.9999% certain to be spam).
You could easily modify a Bayesian filter to have time-decaying weights, but if the change in metaphor leads somebody to come up with a good insight, then perhaps this is useful. Mathematically, though, the equations look very similar.
I concur that the vast majority of people are good and decent. But the situation in Iraq demonstrates that it is possible to mobilize a few thousand people willing to look civilians directly in the face and then kill them. Even in the US, many people every day kill and injure others in the name of money; why not in the name of a political or religious cause?
We slashdotters generally look at security as an all-or-nothing sort of deal: if there's a way around, we assume somebody will find and exploit it. And yet it doesn't appear to be happening in this instance, and I'm not sure why.
Sure, if you wanted to canonicalize one particular dialect. IPA is excellent at capturing the differences between accents, and if you decide to define a single IPA spelling of a word, you'll make all of the other accents invalid. And make it damn hard to spell.
The received pronunciation is being spread through TV, and the differences in accent are becoming less marked over time. My own accent is pretty close to the received one (I grew up in suburban Maryland), but you can find people living 50 miles from me with accents so different I have a hard time understanding them over the phone.
Personally, I think the diversity of accent lends tremendous charm to the language. I'd hate it if everybody spoke the same way.
American Sign Language is considerably different from both written and spoken English, but there's no written equivalent for ASL except written English.
ASL isn't just a word-for-word translation of English. It would be extremely tedious to sign that way; you could more-or-less do it, but you'd sound very stilted (just like if you spoke in the same way you wrote). It's not even just an abbreviation; there are syntactic structures used in ASL that have no exact word-for-word correspondence to either spoken or written English.
We all learn at least two languages, one spoken and one written. They're usually closely related. Orthography is hardly the biggest difference between the two.
Slashdot Mirror
Interesting info. Thanks.
I didn't realize that state compacts had to be approved by Congress. I suppose it makes sense under the Interstate Commerce clause, but I've been unable to find any good cites on this. Can you point me in the right direction?
I'd be curious to see how enforceable the contract turns out to be. I can imagine a state changing its mind midway through the voting, or secretly changing its vote, or something. If the other states sue to enforce the contract, would it prove valid?
It does make recounts rather a mess. One advantage to the electoral college system is that as messy as the Florida recount was, at least it was in only one state. The election of 2000 was very close even in popular terms, and without the electoral college every single state would have ended up having a recount, because every single vote would matter. But gosh, other countries manage to work it out.
The states that have already talked about signing on are big states: California, New York, Colorado, Illinois and Missouri. States who are under-represented in the electoral college. The little states, who currently benefit from having their individual votes be worth nearly 3 times as much as a voter from California or New York, will pitch a major hissy fit.
I haven't run the numbers, but I suspect that such a scheme will tend to favor Democrats over Republicans, at least with the current distributions. Those small states tend to be red states. Certainly the one recent example where one can point to a candidate getting an advantage from the electoral college favored a Republican over a Democrat, so any attempt to swing it towards a proportional vote will be greeted in red states as an attempt to make it more blue.
And like any good programmer, you're willing to share the source for free.
Just don't expect you to maintain it.
Hey, sorry if I offended. I was trying to make a joke; I apologize if it landed with a thud. I have several friends who are LD in a variety of ways, so I sympathize.
(What I could use is a "bad taste" filter on my posts. This is the fourth time in the last month I've stuck my foot in my mouth in public.)
1. The Box Model, is the math 9in? IE finally not backwards from every one else, does it now make sense? Will 'Border' not be full scree when I just set them to '30px'?
2. Oh one thing I am happy about in Fire Fox that is a long time coming for me is the spell check,
We noticed.
Actually, "scree" is a perfectly valid word, and OpenOffice doesn't flag "9in" (or "30px" for that matter). And it's sure not going to tell you that "Firefox" is one word.
But really, you're right: a spell check will probably prevent a lot of typos from getting out into the world. I've had to fix a few of my own by hand in typing this, and nothing kills a "spelling nazi" joke like screwing it up yourself.
I guess what they're getting at is that IE has a button to let you bring up the bookmarks toolbar, rather than having to go to the menu or use a keyboard shortcut.
Or maybe they're referring to the "Add..." button on the Favorites sidebar.
Either way, I'm sure that if anybody cared enough to create this feature it would be an extension. It sounds like the sort of thing you do as part of a tutorial titled "My First Incredibly Easy Extension". It hardly sounds like a feature on par with a pop-up blocker. And in fact given that it's a button I'd never use I'm just as happy that it's not sucking up screen real estate on the navigation toolbar.
It just goes to show you that one person's completely useless screen-clogging waste is the next person's critical can't-live-without feature.
On the font-weight vs. text-decoration distinction, CSS is inheriting the terminology from typesetters and type designers. Technically, "Times Roman" and "Helvetica" designate "typefaces". "Times Roman Bold" and "Helvetica Italic" are "fonts". It's a property of the design itself; the bold and italic aren't simply automatically-derived versions of the typeface but require an artist to sit down and design them separately. (Some even incorporate the size; Times Roman 24 isn't always just a zoomed-out version of Times Roman 12).
(It gets even more complicated with the notion of "font families", but I don't understand the distinction there, either.)
Underlining, on the other hand, is just something you do to it; there isn't any "Times Roman Underlined". That makes it a property of the text, not of the font or face. You don't need a designer to add it.
It sucks that you need such details to do something that you get just by pushing a button in every WYSIWIG word processor in the world. What we need, and what I haven't seen yet, is a WYSIWIG designer for CSS. I envision something equivalent to what Word and OpenOffice call "character styles", but frankly most people don't use them even when they're available.
And Word/OpenOffice still lack (for the most part) an equivalent of CSS layout, which is the part I still find hard. As you point out, CSS's box model seems to be missing some really basic ideas, and that causes many people to just say, "This is 300 pixels wide and it looks fine at a font size I'm comfortable reading and I don't want to f*** with it any more."
In theory, that's why the cable/telcos want to break "net neutrality": they want to charge Apple to ensure the bandwidth.
What drives the people campaigning for net neutrality is the potential for abuse: cutting or drastically throttling traffic from any site who won't pay you, no matter how little traffic they actually generate, or perhaps sites whose politics the ISP disagrees with.
People are buying the TV shows at 320x240 and watching them on their iPods. It stuns me, but the convenience seems to outweigh the resolution issues for a lot of people. For movies you've got the additional hassle of the different screen shape; perhaps the latest round of iPods have a 16x9 screen.
We see exactly the same argument whenever there's a HD article: lots of Slashdotters claiming that 640x480 is plenty for them and they don't feel the need to spend a lot of money on a new TV and player for the higher resolution.
Good luck on your fiancee's iPod.
The article says that they had "not paid federal wagering taxes on $3.3 billion on wagers taken by the firm". At a guess, if the house is taking a 2% cut, at a 30% tax rate, that's almost $20 million that they owe. That's enough money to justify arresting somebody and having them extradited.
(There's a whole big deal going on between the US and Britain with respect to extradition, having to do with levels of proof, but I'll assume for the moment that they've got some proof on this guy. It's a separate argument.)
I've never been exactly sure what "racketeering" is. It's kind of the US-white collar equivalent of what the Soviets used to call "hooliganism": a generic crime applied to people doing scuzzy stuff in bulk where you don't want to (or can't) pin any individual crimes on them.
The charge makes me nervous, because its vagueness makes it subject to abuse. But tax evasion is a crime I an understand. There's also the fraud charges, which is a more specific crime but I don't see specific allegations in the article. Presumably they're in the indictment. It lets them avoid having to test their dubious bans against internet gambling directly.
It's not really a question of "enough". It's a question of what a stock is worth.
A naive interpretation of a stock's worth is that you take the company's assets, sell them, and divide them up among the shareholders. But a company is clearly worth more than that: it makes profits. So you should add to the worth the amount of money you could expect to make from it if you were to divide up the profits.
But what are the profits going to be? That's hard to say, and much magic goes into figuring it out. People get some idea of what they think it's worth, and they price the shares accordingly. Their guesses aren't secrets: the company (sometimes) says what they think the profits will be, and so do analysts, and they share them.
Make less than that, and the share price goes down. It has to: that share is worth less than you thought it was. It doesn't matter if it's still profit; your share of that profit is less than you thought it was going to be, and you'll pay less for it. And if you extrapolate form how much they missed this profit goal to the future ones that you already had figured in when you made your stock price, you'll pay even less.
This is a vast oversimplification, sadly. The price goes down by more than the profit miss indicates because there's an overcorrection: people see others selling the stock and so they figure it's going down and sell theirs, too. Predicting that herd behavior is obnoxious, but if you were to buy-and-hold the stock for a long time the overcorrections will even out over time. Or you can day trade and take advantage of those overcorrections on a daily basis, but if you do your first investment had better be in Tums. In bulk.
In other words: the share price includes a "discount" of what they expect future earnings to be. Miss those earnings, and you have to recompute the discount. An actual decline would be a disaster, but even less-increase affects your calculations.
I believe you're correct. I'm under the impression (and I haven't read the decision) that the judge wasn't exactly encouraging kinderstart to re-file. He was just pointing out to Google that there was still an avenue for this to keep coming back at them, even while he was rejecting kinderstart's original claim of antitrust.
That's correct. Kinderstart would further have to show that a pagerank is some kind of objective fact, not just an opinion. I detailed this in a response to a sister posting to yours, but the gist is that Kinderstart could conceivably demonstrate that a low rank isn't merely "you suck" but "you are worthless", and they'd refute that statement by showing that there is useful information on their site. Google gave them a 0, and they'd argue that connotes "completely worthless", which is closer to an objective statement of fact than merely "you suck".
I doubt it would work, and I suspect the judge does, too, but he didn't want to rule on that argument if it hasn't been made yet. As long as it has a chance of working, he has to permit it to happen.
They'd have to show that Google is more than just saying "You suck". It's perfectly legal to express an opinion. Even if they can prove malice, their next step is to prove that the rank is a false statement of fact.
So the judge is saying that he's also not completely closed to that idea, either. Google would probably say that a page rank is a matter of opinion. They'll respond that it's not; it's an objective number, and you artifically lowered ours. Google will say that the number is a product of an algorithm, and that the algorithm encodes certain opinions that they have, including "You suck."
Or kinderstart could say, "Google makes a statement that our site is less 'good' than site X. We submit a bunch of paperwork to show our site is better. Therefore their statement is false, and we show malice with the artificial page-rank lowering."
I doubt the judge would buy that, either, if for no other reason than that it makes every single Google decision actionable. But he didn't want to rule it out completely with prejudice.
An RFID chip is really, really, really simple. So simple it doesn't have any power on board; it just uses the power of the radio you beam at it. And so simple it costs less than a penny to make. You're never going to get a fingerprint reader on one.
The judge says he's open to the idea that if Google manually changed the pagerank to zero, then maybe it's defamation, which is a valid reason to bring a suit. It's equivalent to saying, out loud and deliberately, "You suck". He's saying that if they do it manually, that could be evidence for malice, which is a requirement for a defamation suit. (At least according to Wikipedia; IANAL.)
I suspect that the judge wanted to leave the idea open, since it hadn't been explored completely and therefore he couldn't absolutely rule it out, but I doubt it would fly. First they have to show that malice, and I can't help but think that they'll have a hard time with that. They'd need somebody with inside knowledge of the decision process; the judge has pretty much said that the lower number is not in and of itself evidence of malice.
I'm sure they've got something they'll throw at this, so I doubt it's the last time we've heard of them. I suspect from here it'll be:
kinderstart: we have evidence
Judge: no, you don't. Go away. You suck.
kinderstart: We sue you! We sue you!
And the great cycle of life begins again.
Even without getting runner's high, I'd rather be out doing something than just lying there getting sweaty.
(I've gotten actual runner's-high euphoria once, perhaps twice. Usually it's just more slogging.)
Glancing at the page, my eyes combined "Net Neutrality" with "Online OSes" on the next line and reported "Net Neuroses" to my brain.
I think that means something, but I'm not quite sure what.
Even the notion that "constants are constant" is just a special case of the assumption that there are fundamental laws, and that they tend to be relatively simple. The constants appear, more or less by magic, in formualas. That magic is a source of wonder, and often the whole formula is replaced by a radically different one, from which the constant falls out as a good approximation from the limited range of experiments we've been able to do so far.
But the new formula usually has constants of its own. There's a non-rational belief that sooner or later we'll reduce it all to one very small set of constants that seem less like magic, but I can't prove it.
To my mind, the difference is in the nature of the principles. A scientist's principles are very short and simple, and pretty much everthing derives from there. They're so simple that they seem pretty much obvious. (Though in general "obvious" is not a good thing; many obvious things are wrong.)
By contrast, religions usually have vastly complicated sets of principles when it comes to the nature of the universe, usually a God with a mind whose decisions can seem arbitrary and capricious, and that same God also imposes a rather complex set of moral principles which cannot be derived from the physical properties of the universe. They often further posit punishments and rewards for those.
Religions believe that the world is a very complex place. Scientists believe that the world is, fundamentally, a place with simple rules that appear complex through interaction. That simplicity is very compelling to me, but I can't really prove it.
Yeah, I noticed the same thing. In one sense it's kind of irritating to have the insinuation perpetrate the myth that scientists have a non-rational belief equivalent to a religious belief, and that these scientsts are some kind of heretics. We know what they meant, but still...
A more precise headline is somewhat harder to write: "Scientists find evidence that they may have to refine or even refactor some really, really well-demonstrated theories" isn't nearly as punchy.
(Scientists do, in fact, have non-rational fundamentally held beliefs, but they're nothing so simple as "Einstein was right, Darwin was right". Trying to convince somebody that a scientist's real religious belief is "The universe has some sort of fundamental, objective, and probably comparatively simple law, one that we can understand or at least produce successively more acurate approximations, one that can be modeled mathematically and is true over all space and time, one that makes predictions that can be tested and will stand up to all such tests all the time" is rather more complicated and less fun. And yes, I recognize that my approximation of that belief above is both more complicated and less accurate than some other formulations, but I'm already drifting dangerously off-topic.)
Ultimately, very little. At core, they're probably identical techniques, and if I were reviewing this as a scientific paper I'd ding them for not answering exactly that question. There are such strong parallels between the two (train them on known data, add up probabilities, cut stuff on a threshold) that I strongly suspect that they're identical.
There are useful things to be gained from a change of metaphor. For example, one difference between this and most bayesian spam filter implementations is that this explicitly incorporates a decay function. That could be useful, if a word that used to be common in spam no longer is (e.g. if I actually decided to buy a Rolex, it's no longer a strong spam indicator, whereas right now any email mentionining "Rolex" is 99.9999% certain to be spam).
You could easily modify a Bayesian filter to have time-decaying weights, but if the change in metaphor leads somebody to come up with a good insight, then perhaps this is useful. Mathematically, though, the equations look very similar.
I concur that the vast majority of people are good and decent. But the situation in Iraq demonstrates that it is possible to mobilize a few thousand people willing to look civilians directly in the face and then kill them. Even in the US, many people every day kill and injure others in the name of money; why not in the name of a political or religious cause?
We slashdotters generally look at security as an all-or-nothing sort of deal: if there's a way around, we assume somebody will find and exploit it. And yet it doesn't appear to be happening in this instance, and I'm not sure why.
Sure, if you wanted to canonicalize one particular dialect. IPA is excellent at capturing the differences between accents, and if you decide to define a single IPA spelling of a word, you'll make all of the other accents invalid. And make it damn hard to spell.
The received pronunciation is being spread through TV, and the differences in accent are becoming less marked over time. My own accent is pretty close to the received one (I grew up in suburban Maryland), but you can find people living 50 miles from me with accents so different I have a hard time understanding them over the phone.
Personally, I think the diversity of accent lends tremendous charm to the language. I'd hate it if everybody spoke the same way.
American Sign Language is considerably different from both written and spoken English, but there's no written equivalent for ASL except written English.
ASL isn't just a word-for-word translation of English. It would be extremely tedious to sign that way; you could more-or-less do it, but you'd sound very stilted (just like if you spoke in the same way you wrote). It's not even just an abbreviation; there are syntactic structures used in ASL that have no exact word-for-word correspondence to either spoken or written English.
We all learn at least two languages, one spoken and one written. They're usually closely related. Orthography is hardly the biggest difference between the two.