We're not building nuclear power stations for one simple reason: We don't know what to do with the waste byproduct yet
I assume you're rejecting the solution presently used by the fossil fuel industry, which is just to dump it directly into the environment at the point of generation, right?
'cause if that's on the table, well, problem solved.
But if you, quite reasonably, reject this solution then it shouldn't be permissible for the fossil fuel industry either. So comparing apples to apples we see that nuclear power is much better off.
The volume (and mass) of waste per kilowatt hour of power is orders of magnitude lower for nuclear than for fossil fuels.
The bulk of nuclear wastes can be cost effectively reprocessed to make more fuel, reducing the amount of new fuel that needs to be mined at the same time as you reduce the amount of wastes that need to be disposed of; neither is the case for fossil fuels.
Much of the remaining nuclear waste material has a short half-life, meaning after a relatively brief period of storage it is no longer dangerous. Not so fossil fuel wastes, which are essentially stable and remain just as dangerous forever.
The remainder of the nuclear waste material is long-half life solids which, due to the very nature of half lives, aren't very radioactive. This means they can be handled with reasonable precautions which is a double win since many of them are economically useful--unlike the waste products of fossil fuel use which are either to valueless (like CO2) or too dilute (like mercury) to be economically recovered.
I have yet to hear a single explanation of why an algorithm implemented with molecules (e.g. chemical process patent) is peachy goodness while an algorithm implemented with bits (e.g. software process patent) is Teh Evil. Every single argument -- every single one -- against algorithm patents apply to chemical process patents, but in practice no one ever makes that case against chemical process patents, which at a minimum raises a lot of questions about the integrity of the position.
You're drawing the analogy at the wrong level. An algorithm, as a ideal entity, with certain useful properties, is analogous not to a chemical process but to a chemical which is also an ideal entity with certain useful properties. Carbon tetrachloride has the same basic properties no mater how you produced it, just as a bubble sort does.
Algorithms are not patentable in many countries. So what people do to patent them is they say that they apply for a patent on a "computer system running the algorithm described". Which is a reasonable thing to do since it's pretty hard to run algorithms on a sheet of paper these days.
The counter attack there being the "not obvious" leg of patentablity. Using a tool for it's intended purpose is considered obvious, and therefore not patentable. Since running algorithms is what computers do, by definition, it's a short step to an "obvious, therefore not patentable" attack.
Basically, the argument is if you have a nail that isn't patentable and a hammer that was specifically designed to hit nails with then hitting the nail with the hammer is obvious and not patentable.
Hmmm. If this were done someplace that was worth the effort (no idea what that city is like) it could potentially be a great way to keep track of where the cops were and maybe even what they were up to.
I'm not disputing your claim, I'm disputing its relevance: the conceptual basis of the two words are identical.
And as for DDosing a guy with a guy, that would be very effective; I don't know how to do it, but if there were some action that could be taken by a group of people to collectively render the gunman ineffective (causing him to miss, or his gun to jam, or whatever) that would be exactly what was needed.
Vigilantism, means, at the root, being vigilant. While it might be nice in theory to sit on your hands and wait for someone else to be vigilant on your behalf, we're doomed as soon as everyone takes that attitude.
If there's a guy in a tower with a machine gun taking shots into the crowd bellow, and some subset of the crowd has the ability to DDos, what would you want them to do?
Episode I scale fail. Let's hope it doesn't evolve into Episode III scale fail.
Star Wars Episode III may suck massively. I can't say for sure, because I have yet to manage to sit through the whole thing. Maybe five minutes in or so I start thinking how much the chancellor reminds me of Joe Lieberman, or remembering how well Natalie Portman acted in Beautiful Girls. By ten minutes I'm reading whatever is lying around or playing with my phone. And then I kind of wander off to defrost the freezer or something.
Thoughtfully, Spears has gone public with her story as a warning to others not to fall victim.
Perhaps. But Occam's razor suggests that it never occurred to her that there might be a downside to publicly admitting to being this stupid, and she went public not "thoughtfully" as a "warning to others" but rather unthinkingly as a further example of what happens when you never think things through.
--MarkusQ
Decompiler vs. Disassembler
on
The IDA Pro Book
·
· Score: 3, Insightful
Note that there is a distinction between dissassembler and decompiler - you seem to be describing it as a decompiler, which it is not.
*laugh* I think what's really going on is that I bounce between levels so much that I don't really honor the distinction. Asm, HLA, LLL, HLA, scripting languages, TILs, SPILs, DSLs, it all kind of blurs together if you step away from it just a short distance. Not that I don't see such distinctions, just that I don't always see them in the traditional places. For instance, I see a much bigger divide between pairs like Haskel v. C or SQL v. prolog than I do between C and Assembly.
But yes, I see your point, for people who aren't comfortable reading Assembly and expect it to give them C++ or something the distinction would be important. Effectively, the choice of source language is one of the things that gets lost along with procedure names, module structure, and the like. You may be able to infer it but IDA isn't going to hand it to you and there is no certainty you'll be correct.
--MarkusQ
IDA is a dissassembler
on
The IDA Pro Book
·
· Score: 5, Informative
Since the review doesn't really make it clear, IDA is a dissassembler. It allows you to take an executable for which you don't have source and construct a (generally partial) representation of what a program that would have produced that executable would look like. It can't of course give you back the actual source code (comments, variable names, etc. being lost forever) but it gives you a much, much better idea of what's going on than a hex dump would.
accelerating fuel forward so you can spit it back later.
I have no idea what that even means, or is even supposed to mean. I quote it only to highlight that the source of your skepticism seems based entirely on a gross misunderstanding of the technology involved.
I'm not the person to whom you were responding but I suspect the misunderstanding is on your end, not his. The meaning of the phrase is quite clear; in a system with sustained thrust the fuel (and reaction mass) used in a later portion of the trip has to be accelerated (along with the rest of the ship) for the whole proceeding portion of the trip. This means that, early in a long trip, the majority of the fuel/reaction mass you use accelerating the remainder, and only a small fraction is accelerating the payload. That's why large rocket use stages.
The other advantage is maximum top speed. If your hydrazine rocket can expel mass at, say, 1000 mph (making numbers up here) then the top speed of your rocket is 1000mph for reasons I hope are obvious.
The "reasons" may be obvious to you, but they aren't valid. The actual relationship between final speed (from a standing start in some reference frame) and the exhaust velocity has as a factor the natural log of the starting mass over the payload mass. So (to use your made up numbers) if you started with a ship that was 90% hydrazine (by mass) your final velocity would be 1000*ln(100/10) mph or about 2300 mph, over twice your exhaust velocity. If the ship was 99% fuel, the final velocity would be 4600 mph, and so on.
So, just as the Earth enters a significant cooling trend (~20-30 years of cooling, at least) we should consider something to make it cool even faster?
You do realize, don't you, that there is absolutely zero evidence to support this "trend" you refer to? That it is an entirely hypothetical construct, based on little more than numerology? Yes, there was a sunspot minimum during the European little ice age, but there is ample reason to believe that there was no cause and effect relationship: cooling was regional, not global, it was strongly correlated with a period of increased volcanic activity (which, unlike the sunspot theory, has a clear, well demonstrated causative correlation), and the solar constant (the amount of energy we receive from the sun) did not change significantly.
It's my opinion, after considerable research, that we don't yet know enough to make long-term climate predictions, MUCH LESS BASE POLICY ON THOSE PREDICTIONS!
Ah, the irony. "We" don't know enough to make long term predictions after "considerable research" but somehow you can tell us that a 20-30 year cooling trend is coming based on...what?
Let's see, somebody already got the "tiring" quip and the "40,000 curies is a lot" nonsense, so I'll take a few of the points the other responders left.
The hydrogen explosion was only a small one - well before the large amount of hydrogen accumulated from the reaction between the hot zirconium fuel rod cladding and the cooling water. The thing that kept it from being devastating was that conditions changed where it didn't ignite.
The "conditions that changed" were that all the O2 was used up in the explosion. So there was no more to react with the remaining H2, and so it couldn't possibly explode. H2 in such an environment is essentially an inert gas.
I'll leave your claim that, as radioactive materials decay, "you get more radioactivity growing into your sample making it more radioactive than it originally was" alone in case someone else wants to play. (Hint: if that were true, and given that the Earth's crust was rather radioactive 4 billion years or so ago, what should we expect to see today?)
Uh, it really is dangerous. That's why nuclear power plants are considered terrorist targets.
Be suspicious of any chain of reasoning based on taking what Homeland Security et al think as true. These people are fear mongers, and use artificially created fear to control the masses. For some reason their otherwise incoherent policies always seem to align on one point: they increase the profits of big oil.
While you are correct that Chernobyl was a bad design and an ill-conceived experiment started the disaster, do you recall what caused Three Mile Island or what the consequences might have been had the hydrogen bubble ignited?
Oh come on. The hydrogen did ignite and nothing significant happened. The "bubble" was what was left over after all the hydrogen that could have burned already had. The public and the environment suffered no injury, and the whole thing was blown way out of proportion.
And the bigger problem is the cost and various issues with properly sequestering the waste.
It's only a problem because we have been hornswaggled into thinking of it as "waste" instead of thinking of it as "fuel reserves." If you want to suppress any technology try this simple trick:
Convince people that some intermediate product of the system is "waste"
Convince them that the only thing they can do with it is store it
Point out the logical consequences of these absurd assumptions.
For example, if you could convince people that they could only use 10% of the gas they put in their cars and had to save the other 90% forever, what would happen to the auto industry?
Incidentally, the whole "longer half-life == more dangerous" talking point is stupid. Saying that something has an enormously long half-life is just another way of saying that it is relatively stable. It's the things with the short half-lives you need to worry about. The tungsten in your lightbulbs, for instance, has a half-life of around 1,000,000,000,000,000 years. Buy the backwards logic of the anti-nuke people, it should be terrifying stuff, much more dangerous than uranium with a half-life of only 100,000 to 1,000,000,000 years, right?
-- MarkusQ
Yes, I agree. I hadn't heard of the median/mode weighting that you mention as giving the best fit. Do you know a reference? Thanks.
No, it was a heuristic I picked up from a statistician I worked with years ago, and I'm not even certain I've stated it correctly.
The basic idea is a generalization of the rounding rule. If you have a bunch of values evenly distributed between 0.0 and 1.0 you can approximate their sum by counting how many are >= 5.0 and multiplying by 10. So I see already that I got it wrong. The real rule would be count everything over the median as twice the mean, and the mode doesn't figure into it. Unless I'm still remembering incorrectly.
That should teach me to post late at night. My advice would be, if you need to know the actual rule, consult a statistician or a good statistics book, not random programmers on slashdot.
--MarkusQ
P.S. My failure to remember the best of many bad ways to doubly aggregate data actually underscores my main point, which is that its a very tricky thing to get right, and even in the best of circumstances you have to be careful drawing conclusions from the results.
* And if
Correct, but bear in mind that this isn't a measure of sea ice coverage either; it's a doubly aggregated value that must be taken with a dram of salt water. While it is true that the percentage of sea surface covered with ice is strongly correlated with local albedo, the percentage of sea surface parcels which are covered with more than a certain threshold percentage of ice is much less strongly correlated, especially when that percentage is far from the median. IIRC correctly, using the median value as the threshold and weighing it as the mode gives the best fit.
To but this more concretely, a set of parcels with ice coverage of (0.16,0.16,0.16,0.16,0.16) would have an albedo of 0.84*sea + 0.16*ice, but show up in this survey as 100% ice. A set with (0.14,0.14,0.99,0.14,0.14) would have an actual albedo almost twice as high (0.69*sea + 0.31*ice) but show up in this survey as only 20% ice.
It's important to keep in mind that this isn't a measure of how much ice there is in the arctic.
The figures they are reporting are sea ice coverage estimates, and typically work as follows: the arctic is broken up into a grid, and for each area of the grid which does not fall on land they ask the question "is >15% of the surface covered with ice?"
If the answer is yes, it's counted as "ice;" if not, not.
There are several ways this can give results you wouldn't expect:
If one cell of the grid is 85% ice covered and the eight adjacent cells are ice free, this counts as 1/9 of that area being ice. If some of the ice melts and the rest disburses so that the original cell and four neighbors are now all 16% ice covered, it counts as five times as much ice coverage (5/9) even though the total amount of ice went down
For you, someone telling you "Yes, she used private email illegally for government business" might suffice - but let's put the shoe on the other foot - taking someone else's word for something like this, when you don't know THEIR agenda, is usually only done by fools - or by people who WANT to believe. That makes sense, right?
Yes, it makes sense though I think you are carrying it too far. Demanding to see an actual email in violation is a pretty silly standard since a person just a shade more skeptical could simply say "ah, but they were forged!"
A better standard (IMHO) is to take a mix of occam's razor with trust in our judicial system. If a judge or jury, presented with more evidence than I have time or even authority to examine, comes a conclusion on a point of fact, that is a reasonable starting point. Yes, it could be wrong, but odds are it isn't.
In the present case, the combination of circumstantial evidence, sworn testimony, and partial rulings, and the internal logic of the situation does not support the assumption that she was on the up-and-up. If you find that the local convenience store is keeping two sets of books, has frequent unexplained visits from people who just chat with the clerk but never buy anything, you might suspect that something was up. If you find out that several ex-employees have been called to testify before a grand jury and a judge has slapped the owners for obstruction, a reasonable person might well conclude that a crime or crimes have been committed without needing to see formal proof.
I'm not saying you'd always be right, but that sure is the way to bet.
I'm still waiting for something besides hearsay, really.
And you are free to do so. The sort of evidence that would convince a reasonable person is never sufficient to convince a determined skeptic (or, I suspect in this case, a true believer in the contrary position). Sworn testimony of people in position to know, the finding of independent investigators, and common sense are sufficient to convince me pending conflicting information, and I suspect they sufice for most people. But you are correct, there is no way to know for sure in a case like this without being Palin herself since any "evidence" could be the result of a conspiracy to discredit her. Maybe her aides are in on it. Heck, she could be in on it too, so even if she admitted it*, we wouldn't really know, would we?
So just for giggles, let's turn the question around. Do you have any proof she never used an external email account to conduct state business? Or are you just choosing to adhere to the less likely position because it's more compatible with your political views?
--MarkusQ
* Given the way she changes her answers and often seems to just say whatever pops into her head, it's not impossible that she might come out and admit it.
You did notice, of course, that these emails were from her official government account, right?
Not all of them, at least according to the headers and response chains in the mails that were released (from other people's e-mail accounts). Now before you say it yes, it is possible that her assistants and her husband (or rather, someone pretending to be her husband, since actually including her husband on some of the e-mails would in and of itself have been illegal) were all part of a plot to create the impression that official e-mails went through a non-official account, by forging message headers and such, but occam's razor say she was just doing something she shouldn't have been, got caught, and is trying to cover it up.
Just a thought, and a bit off-topic, I know, but I was wondering if there is an Absolute Maximum Temperature?
In at least one sense, there obviously is a highest possible temperature, and in another, there can't possibly be. If there is a highest temperature, it is probably the planck temperature, unless it's the hagedorn temperature, or, under a certain crafty merger of cosmology and negative temperatures the maximum might be -0k.
Unless it's something else. Or unless there isn't one.
Ah, my mistake. When I read your list it seemed you were going for the last option as the obvious one. Seeing as how this object's density is pretty much exactly in line with what you'd expect from a brown dwarf, I should have realized that you actually meant that one.
No, you were right in your interpretation, I was mostly betting on data error, though brown dwarf would be a good second bet. I'm going off a physics education that's got twenty plus years of dust on it and half these posts have been made with a kid or a cat on my lap, so I won't claim rigorous analysis here, but I'm not convinced. There are various reasons (tidal interactions, catchment competition, etc.) that I'm unconvinced that it could have formed in situ. And there are statistical reasons for doubting that it migrated there. So I'm skeptical.
I will admit that I've become less so after earlier reports that had the star much denser (based on a mass of 3.7 M(s) vs. the measured 1.37 M(s)) have been corrected.
Density isn't exactly dimensionless, but if you set things up so the density of water is 1 in a system of measurements, the densities of other things (i.e. Lead, Iridium, or this planet) will come out the same numbers, regardless of the units used. So it's not necessary to really specify the units, just that H2O at STP = 1 in whatever system you are using.
That is the sort of unlabeled density I was throwing around in the first place, but IIRC there are some systems in which it truly dimensionless. To get there you start measuring things that are equivalent in the same units. Say you decide to start with meters. You can measure time in meters using the space time formulation (so you are measuring the interval between events via I = sqrt(delta:x^2 + delta:y^2 + delta:z^2 - delta:t^2) instead of measuring just time or just distance). When you do this some things (such as scalar velocity) become unitless ratios. In the case of velocity, they become a number between 0 (rest) and 1 (c). All normal speeds are very small numbers in this system.
Likewise you can measure mass and energy in the same units (e.g. electron volts), rendering things like the power of an explosive or the heat of crystallization of some material as unitless (and very small) number.
When you drag in thermodynamics (IIRC you also have to pull information theory and QM to make this work) you can start measuring even lore things things in the same terms (e.g. information in m^2, where one bit = 1/4 of a square planck length or something like that). Eventually, you get to measuring mass in the same terms as volume, making density unitless.
IIRC, in such a system the density of lead is something like 10^240, making it all but useless for the sort of back-of-the-email-envelope calculation I was doing, but still, it's there.
--MarkusQ
P.S. Although this may seem like a strange way to look at units, remember that we already do some unification--we measure differences in altitude in the same units that we use for north-south and east-west distances, we measure the weight of wheat in the same units we use for rice, and so on. It's just a question of where you stop.
Slashdot Mirror
I assume you're rejecting the solution presently used by the fossil fuel industry, which is just to dump it directly into the environment at the point of generation, right?
'cause if that's on the table, well, problem solved.
But if you, quite reasonably, reject this solution then it shouldn't be permissible for the fossil fuel industry either. So comparing apples to apples we see that nuclear power is much better off.
--MarkusQ
You're drawing the analogy at the wrong level. An algorithm, as a ideal entity, with certain useful properties, is analogous not to a chemical process but to a chemical which is also an ideal entity with certain useful properties. Carbon tetrachloride has the same basic properties no mater how you produced it, just as a bubble sort does.
-- MarkusQ
The counter attack there being the "not obvious" leg of patentablity. Using a tool for it's intended purpose is considered obvious, and therefore not patentable. Since running algorithms is what computers do, by definition, it's a short step to an "obvious, therefore not patentable" attack.
Basically, the argument is if you have a nail that isn't patentable and a hammer that was specifically designed to hit nails with then hitting the nail with the hammer is obvious and not patentable.
--MarkusQ
Ok, security-thinking time...
Hmmm. If this were done someplace that was worth the effort (no idea what that city is like) it could potentially be a great way to keep track of where the cops were and maybe even what they were up to.
--MarkusQ
I'm not disputing your claim, I'm disputing its relevance: the conceptual basis of the two words are identical.
And as for DDosing a guy with a guy, that would be very effective; I don't know how to do it, but if there were some action that could be taken by a group of people to collectively render the gunman ineffective (causing him to miss, or his gun to jam, or whatever) that would be exactly what was needed.
--MarkusQ
Vigilantism, means, at the root, being vigilant. While it might be nice in theory to sit on your hands and wait for someone else to be vigilant on your behalf, we're doomed as soon as everyone takes that attitude.
If there's a guy in a tower with a machine gun taking shots into the crowd bellow, and some subset of the crowd has the ability to DDos, what would you want them to do?
--MarkusQ
Star Wars Episode III may suck massively. I can't say for sure, because I have yet to manage to sit through the whole thing. Maybe five minutes in or so I start thinking how much the chancellor reminds me of Joe Lieberman, or remembering how well Natalie Portman acted in Beautiful Girls. By ten minutes I'm reading whatever is lying around or playing with my phone. And then I kind of wander off to defrost the freezer or something.
--MarkusQ
Perhaps. But Occam's razor suggests that it never occurred to her that there might be a downside to publicly admitting to being this stupid, and she went public not "thoughtfully" as a "warning to others" but rather unthinkingly as a further example of what happens when you never think things through.
--MarkusQ
*laugh* I think what's really going on is that I bounce between levels so much that I don't really honor the distinction. Asm, HLA, LLL, HLA, scripting languages, TILs, SPILs, DSLs, it all kind of blurs together if you step away from it just a short distance. Not that I don't see such distinctions, just that I don't always see them in the traditional places. For instance, I see a much bigger divide between pairs like Haskel v. C or SQL v. prolog than I do between C and Assembly.
But yes, I see your point, for people who aren't comfortable reading Assembly and expect it to give them C++ or something the distinction would be important. Effectively, the choice of source language is one of the things that gets lost along with procedure names, module structure, and the like. You may be able to infer it but IDA isn't going to hand it to you and there is no certainty you'll be correct.
--MarkusQ
Since the review doesn't really make it clear, IDA is a dissassembler. It allows you to take an executable for which you don't have source and construct a (generally partial) representation of what a program that would have produced that executable would look like. It can't of course give you back the actual source code (comments, variable names, etc. being lost forever) but it gives you a much, much better idea of what's going on than a hex dump would.
--MarkusQ
I'm not the person to whom you were responding but I suspect the misunderstanding is on your end, not his. The meaning of the phrase is quite clear; in a system with sustained thrust the fuel (and reaction mass) used in a later portion of the trip has to be accelerated (along with the rest of the ship) for the whole proceeding portion of the trip. This means that, early in a long trip, the majority of the fuel/reaction mass you use accelerating the remainder, and only a small fraction is accelerating the payload. That's why large rocket use stages.
The "reasons" may be obvious to you, but they aren't valid. The actual relationship between final speed (from a standing start in some reference frame) and the exhaust velocity has as a factor the natural log of the starting mass over the payload mass. So (to use your made up numbers) if you started with a ship that was 90% hydrazine (by mass) your final velocity would be 1000*ln(100/10) mph or about 2300 mph, over twice your exhaust velocity. If the ship was 99% fuel, the final velocity would be 4600 mph, and so on.
--MarkusQ
Spot on. Dead on target and a car analogy. You rock.
--MarkusQ
You do realize, don't you, that there is absolutely zero evidence to support this "trend" you refer to? That it is an entirely hypothetical construct, based on little more than numerology? Yes, there was a sunspot minimum during the European little ice age, but there is ample reason to believe that there was no cause and effect relationship: cooling was regional, not global, it was strongly correlated with a period of increased volcanic activity (which, unlike the sunspot theory, has a clear, well demonstrated causative correlation), and the solar constant (the amount of energy we receive from the sun) did not change significantly.
Ah, the irony. "We" don't know enough to make long term predictions after "considerable research" but somehow you can tell us that a 20-30 year cooling trend is coming based on...what?
--MarkusQ
Let's see, somebody already got the "tiring" quip and the "40,000 curies is a lot" nonsense, so I'll take a few of the points the other responders left.
The "conditions that changed" were that all the O2 was used up in the explosion. So there was no more to react with the remaining H2, and so it couldn't possibly explode. H2 in such an environment is essentially an inert gas.
Also, airliners are no threat to containment buildings.
I'll leave your claim that, as radioactive materials decay, "you get more radioactivity growing into your sample making it more radioactive than it originally was" alone in case someone else wants to play. (Hint: if that were true, and given that the Earth's crust was rather radioactive 4 billion years or so ago, what should we expect to see today?)
--MarkusQ
Be suspicious of any chain of reasoning based on taking what Homeland Security et al think as true. These people are fear mongers, and use artificially created fear to control the masses. For some reason their otherwise incoherent policies always seem to align on one point: they increase the profits of big oil.
Oh come on. The hydrogen did ignite and nothing significant happened. The "bubble" was what was left over after all the hydrogen that could have burned already had. The public and the environment suffered no injury, and the whole thing was blown way out of proportion.
It's only a problem because we have been hornswaggled into thinking of it as "waste" instead of thinking of it as "fuel reserves." If you want to suppress any technology try this simple trick:
For example, if you could convince people that they could only use 10% of the gas they put in their cars and had to save the other 90% forever, what would happen to the auto industry?
Incidentally, the whole "longer half-life == more dangerous" talking point is stupid. Saying that something has an enormously long half-life is just another way of saying that it is relatively stable. It's the things with the short half-lives you need to worry about. The tungsten in your lightbulbs, for instance, has a half-life of around 1,000,000,000,000,000 years. Buy the backwards logic of the anti-nuke people, it should be terrifying stuff, much more dangerous than uranium with a half-life of only 100,000 to 1,000,000,000 years, right? -- MarkusQ
No, it was a heuristic I picked up from a statistician I worked with years ago, and I'm not even certain I've stated it correctly.
The basic idea is a generalization of the rounding rule. If you have a bunch of values evenly distributed between 0.0 and 1.0 you can approximate their sum by counting how many are >= 5.0 and multiplying by 10. So I see already that I got it wrong. The real rule would be count everything over the median as twice the mean, and the mode doesn't figure into it. Unless I'm still remembering incorrectly.
That should teach me to post late at night. My advice would be, if you need to know the actual rule, consult a statistician or a good statistics book, not random programmers on slashdot.
--MarkusQ
P.S. My failure to remember the best of many bad ways to doubly aggregate data actually underscores my main point, which is that its a very tricky thing to get right, and even in the best of circumstances you have to be careful drawing conclusions from the results. * And if
Correct, but bear in mind that this isn't a measure of sea ice coverage either; it's a doubly aggregated value that must be taken with a dram of salt water. While it is true that the percentage of sea surface covered with ice is strongly correlated with local albedo, the percentage of sea surface parcels which are covered with more than a certain threshold percentage of ice is much less strongly correlated, especially when that percentage is far from the median. IIRC correctly, using the median value as the threshold and weighing it as the mode gives the best fit.
To but this more concretely, a set of parcels with ice coverage of (0.16,0.16,0.16,0.16,0.16) would have an albedo of 0.84*sea + 0.16*ice, but show up in this survey as 100% ice. A set with (0.14,0.14,0.99,0.14,0.14) would have an actual albedo almost twice as high (0.69*sea + 0.31*ice) but show up in this survey as only 20% ice.
--MarkusQ
It's important to keep in mind that this isn't a measure of how much ice there is in the arctic.
The figures they are reporting are sea ice coverage estimates, and typically work as follows: the arctic is broken up into a grid, and for each area of the grid which does not fall on land they ask the question "is >15% of the surface covered with ice?"
If the answer is yes, it's counted as "ice;" if not, not.
There are several ways this can give results you wouldn't expect:
--MarkusQ
Yes, it makes sense though I think you are carrying it too far. Demanding to see an actual email in violation is a pretty silly standard since a person just a shade more skeptical could simply say "ah, but they were forged!"
A better standard (IMHO) is to take a mix of occam's razor with trust in our judicial system. If a judge or jury, presented with more evidence than I have time or even authority to examine, comes a conclusion on a point of fact, that is a reasonable starting point. Yes, it could be wrong, but odds are it isn't.
In the present case, the combination of circumstantial evidence, sworn testimony, and partial rulings, and the internal logic of the situation does not support the assumption that she was on the up-and-up. If you find that the local convenience store is keeping two sets of books, has frequent unexplained visits from people who just chat with the clerk but never buy anything, you might suspect that something was up. If you find out that several ex-employees have been called to testify before a grand jury and a judge has slapped the owners for obstruction, a reasonable person might well conclude that a crime or crimes have been committed without needing to see formal proof.
I'm not saying you'd always be right, but that sure is the way to bet.
--MarkusQ
And you are free to do so. The sort of evidence that would convince a reasonable person is never sufficient to convince a determined skeptic (or, I suspect in this case, a true believer in the contrary position). Sworn testimony of people in position to know, the finding of independent investigators, and common sense are sufficient to convince me pending conflicting information, and I suspect they sufice for most people. But you are correct, there is no way to know for sure in a case like this without being Palin herself since any "evidence" could be the result of a conspiracy to discredit her. Maybe her aides are in on it. Heck, she could be in on it too, so even if she admitted it*, we wouldn't really know, would we?
So just for giggles, let's turn the question around. Do you have any proof she never used an external email account to conduct state business? Or are you just choosing to adhere to the less likely position because it's more compatible with your political views?
--MarkusQ
* Given the way she changes her answers and often seems to just say whatever pops into her head, it's not impossible that she might come out and admit it.
Not all of them, at least according to the headers and response chains in the mails that were released (from other people's e-mail accounts). Now before you say it yes, it is possible that her assistants and her husband (or rather, someone pretending to be her husband, since actually including her husband on some of the e-mails would in and of itself have been illegal) were all part of a plot to create the impression that official e-mails went through a non-official account, by forging message headers and such, but occam's razor say she was just doing something she shouldn't have been, got caught, and is trying to cover it up.
--MarkusQ
Fox "News" already does this when they're running stories about reporters from other news outlets.
--MarkusQ
In at least one sense, there obviously is a highest possible temperature, and in another, there can't possibly be. If there is a highest temperature, it is probably the planck temperature, unless it's the hagedorn temperature, or, under a certain crafty merger of cosmology and negative temperatures the maximum might be -0k.
Unless it's something else. Or unless there isn't one.
--MarkusQ
No, you were right in your interpretation, I was mostly betting on data error, though brown dwarf would be a good second bet. I'm going off a physics education that's got twenty plus years of dust on it and half these posts have been made with a kid or a cat on my lap, so I won't claim rigorous analysis here, but I'm not convinced. There are various reasons (tidal interactions, catchment competition, etc.) that I'm unconvinced that it could have formed in situ. And there are statistical reasons for doubting that it migrated there. So I'm skeptical.
I will admit that I've become less so after earlier reports that had the star much denser (based on a mass of 3.7 M(s) vs. the measured 1.37 M(s)) have been corrected.
--MarkusQ
That is the sort of unlabeled density I was throwing around in the first place, but IIRC there are some systems in which it truly dimensionless. To get there you start measuring things that are equivalent in the same units. Say you decide to start with meters. You can measure time in meters using the space time formulation (so you are measuring the interval between events via I = sqrt(delta:x^2 + delta:y^2 + delta:z^2 - delta:t^2) instead of measuring just time or just distance). When you do this some things (such as scalar velocity) become unitless ratios. In the case of velocity, they become a number between 0 (rest) and 1 (c). All normal speeds are very small numbers in this system.
Likewise you can measure mass and energy in the same units (e.g. electron volts), rendering things like the power of an explosive or the heat of crystallization of some material as unitless (and very small) number.
When you drag in thermodynamics (IIRC you also have to pull information theory and QM to make this work) you can start measuring even lore things things in the same terms (e.g. information in m^2, where one bit = 1/4 of a square planck length or something like that). Eventually, you get to measuring mass in the same terms as volume, making density unitless.
IIRC, in such a system the density of lead is something like 10^240, making it all but useless for the sort of back-of-the-email-envelope calculation I was doing, but still, it's there.
--MarkusQ
P.S. Although this may seem like a strange way to look at units, remember that we already do some unification--we measure differences in altitude in the same units that we use for north-south and east-west distances, we measure the weight of wheat in the same units we use for rice, and so on. It's just a question of where you stop.