Much more interesting (to me) is that the aircraft used (a DeHavilland DHC-3T) is a flying boat built in 1953. Looks great for fishing trips, but they were taking it through dangerous mountains - known for doing unmentionable things to aircraft - during a severe storm. Hands up all those who would want to be in the aircraft shown in the posting under those kinds of conditions. I feel certain that former Sen. Stevens has been in enough light aircraft (Alaska is big and the roads aren't) to know what you can and cannot do. What I cannot fathom is why he, with his knowledge and experience, would take that kind of a pointless, stupid risk.
As for O'Keefe, for all his time at NASA, I doubt he's enough time in aircraft of this vintage and size to know the risks. Being head honcho of an aerospace organization doesn't endow you with the kind of skill and knowledge needed. Nonetheless, he too should have been wary of flying in those conditions in an antique.
Oh, much worse. But let's face it, the viewers are divided into two camps - those wanting spectacular disasters and those wanting spectacular and strategic clothing failures on female crew. Language is not a particular barrier to either of these...
You'll have to ask Arthur Scargill about the first of those, and the Forestry Commission for the latter. (The modern word "forest" actually started in England.) I'm fairly sure that when most of these numbers were drawn up, most UK mines were still open. That would likely mean the UK had as many - or more - coalminers than the US per capita. For lumberjacks, much the same argument applies. The amount of lumber produced in the UK at that time, per capita, would have been very comparable. (You must remember that the US figures are for the WHOLE of the US. Not many lumberjacks in Montana or North Dakota, but there are people. Well, in the case of North Dakota, person.)
I have no idea what +1, Interesting smallpox is. Presumably, it's a mutation of smallpox that has evolved the ability to post on Slashdot. In which case, we're safe. It can't spread. What's it going to reproduce with?
There's that and the difficulty in keeping a camera lens clean when the water spray freezes on impact. Current conditions (as of 10:47pm GMT) look pretty gentle - 3' waves are nothing. I'm fairly sure that rough seas would be nearer an average of 27-30'. For freak waves, 55.8' waves have been observed by the Netherlands lifeboat association. In the North Sea off Scotland, waves have been recorded as high as 61' for multiple and 95' for an individual wave. I don't know if the North Sea gets more of them than other regions, but Discovery might want to consider that NOBODY has ever done a "reality TV show" that included freak waves, and only one or two have ever been filmed at all. Ever. And by far cheaper equipment than Discovery can get hold of. As for the crew's extreme improbability of surviving, it's television! Besides, the ratings would surge like anything, as all the freaks out there wait for the next true-life disaster.
In Netrek, the term is Ogging, after the greatest maniac ever to have played the game. (No, not me. I'm not even close.) There, they'll Ogg anything - space stations especially.
Don't knock those Second Life home owners. So long as their real-estate still has that value in the real world, those owners have the only property in the world the banks can't touch.
I wonder if the fan site has asked the marketing people to intervene. Usually, although lawyers can get lots of money, marketing people HAVE lots of money and that can make a difference to the upper echelons.
Alaskan crab fishing is ok as a "dangerous reality TV" show. Apparently the stats for Alaskan crab fishing is 356/100,000/year. That's a lot - getting on for 1 in every 100,000 per day. (US National Average workforce fatality rate is 7.0/100,000/year.) I wondered if North Sea fishing was worse - it has a vicious reputation and the North Sea has no landmass between it and the north pole. However, statistics indicates that the mortality rate is 151 x national average in the UK, and the UK's national average is 0.5 deaths per 100,000 people. That puts the North Sea fishermen at a paltry 76/100,000/year. Not safe, by any standards, but many times safer in absolute terms. In relative terms, the US' workforce fatality rate is 14x worse than that in the UK, but the Alaskan crab fishing is only 4.7x as deadly as North Sea fishing. By this standard, North Sea fishing is the deadliest fishing occupation relative to the health and safety of the country involved.
A valid point, but making sure things will work in the US should never dissuade anyone from looking at what works elsewhere -- especially if that includes the important elements of studying WHY it works there and what flaws were discovered when the method was implemented. ("Why" has to include not just the day-to-day why but also the pre-conditions that made it workable to begin with.) If all this is known and compatible pre-conditions exist in the US, then it should be possible to make the tweaks needed to adjust for any differences and to prevent the flaws discovered elsewhere.
There is nothing to stop one country learning from another. Learning is good. What is bad is when one country assumes either that what they have learned MUST apply - undiluted and unmodified - or must NOT apply.
Two or three. Since the post can't drop below -1, you only need to bring it up to +1 or +2 to be visible at most levels of filtering. On EARLIER versions of Slashdot, you could indeed bury a post permanently by taking it to -2. You had one person mod up by 1, the rest of your clique then modded down to get the post to -1, then the person who modded up posted, eliminating the +1. This took the post to -2, which is never visible and can therefore never be modded back up (except by sysops). This bug was fixed some time back.
A related quirk was that you could also get to the dizzying heights of +6 by the same method. The total number of +6-modded posts was extremely small, but they did exist. They also caught attention merely because every poster on the site knew damn well that should be impossible.
I agree, but formal education and the Internet (or television) are not mutually exclusive. In the UK, Open University is an extremely old (for mass communication, at least) and highly respected approach, albeit not as respected as the higher-end bricks-and-mortar facilities. Regardless of the fact that the lectures have been distributed for free for, what, 40? 50? years, there has been no evidence in the UK of the kind of shift Bill Gates has suggested. It would seem to follow that merely being free and widely circulated is insufficient. Further, UK citizens have, in general, had far more acceptance for this kind of system than the US - which is why PBS is in such poor shape and educational channels in general in the US are a bit of a disaster. (If Discovery can only afford to show 6 programs a quarter, at their size, and record even less than that, when the gameshow channels can afford to splurge on prizes bigger than the cost of recording a documentary every hour, you can tell something about what the priorities are.)
The answer is maybe. My thought experiment would suggest it may be possible for one possible outcome to alter the state of the universe in which a different outcome happened, which is similar to your own idea. If that is true, then the answer is necessarily yes. There may be other circumstances in which it could happen, but once you allow - even under the most restrictive of circumstances - a Y-piece to be added to the possible outcomes, then it becomes possible for a particle to go through one slit but register as though it went through the other.
To complicate things further, you can do diffraction over time as well as space. I found the original paper and twofollowups but Physical Review is subscription only. If it is possible to diffract in time as well as space, then it automatically follows that events don't just happen and then go away. The different points in time have to be interacting in some way. I do not fully understand the implications of that, but I would interpret it as meaning that it does indeed mean that it is possible to snake between the different possible worlds, or at least some subset of them.
If it was plain, then it would be more obvious to the rest of us. Since not a single one of the replies to me agreed on the issue - with me or each other - it is clear it is not plain. I will certainly look the guy up, but I sincerely doubt it will change my mind on whether the social sciences are currently or could potentially be a science.
Personally I agree with you, and if it's now pretty much solidly known then that's in line with what I'd have expected. My use of Penrose's claim is because I'm extremely wary of asserting that I'm right and a highly respected professor is wrong. (Wary, but I sometimes do that anyway, and I'm right often enough to make me wonder why they're the one with the cool title and tenure.)
I am extremely glad for your point, though, as it does help clarify things.
First, open source code is - and will be - used to kill people, whether you like it or not. If you don't contribute to such projects for moral reasons, you're not really helping those who use the software for moral purposes and you're not really hurting those who use it for immoral purposes.
Second, this is really a question of the level of indirection. If you contribute a patch to a kernel module which can be used by Linux and that version of Linux is downloaded and used by a third-part vendor which supplies the package to the DoD which, in turn, supplies it to the front-line, you've five levels of indirection. If we assume that each level merely doubled the number of people involved, then you would be 1/(2^5)th responsible for the patch being where it was. That would not make you 1/(2^5)th responsible on how it was used, though. You should really weight for that, putting it something closer to 1/(2^(2^5))th. In practice, given the size of the military, the size of the Linux userbase and the size of the Kernel development team, the dilution is closer to 128 rather than 2. So you're looking at 1/(2^(128^5))th of a share in the responsibility.
True, it is not zero, and I do have a serious issue with those who assume that dilution/indirection == zero responsibility, but at that kind of level it's so damn close to zero that it makes no serious odds. If you were to supply that same patch directly TO the front line, specifically knowing its intended use (or specifically not asking), that would be a different kettle of fish. There would be no dilution and no indirection.
Now, working with a project that the CIA probably will never use (they're a political organization, you think they'd use something if it would look like backing down?) and the military in general probably won't use (partly the Not Invented Here, partly because this is ultimately GOTS and the military are strictly prohibited from using GOTS over COTS, hence the absurd number of contractors and the absurd contract regulations and licensing issues) - meh. There's likely far more risk of death and destruction coming from using one of the NASA CFD packages and handing back a patch (as that might well be used for designing military aircraft).
This piece of software might actually be quite handy for the open source community, though. And the encryption community. There's lots of flamewars over stupid issues. The current one over on the SHA-3 mailing list is over whether NIST's security requirements should be met or if designers can "cheat" if nobody is likely to break the algorithm anyway and to be "secure" means to be slow. To me, this is stupid. You use algorithms that are "good enough" where you need the speed and you use algorithms that are "secure" when you need the security. Try getting that through some incredibly thick craniums some time. However, if a piece of software produced exactly the same advice, I can bet you anything that those same people would pay attention.
The same goes for whether Linux audio should be robust or realtime. GUIs should be KDE or Gnome. SELinux or GRSecurity. Logging FS versus Journaling. We'd be miles further along if the petty bullshit of one-upmanship didn't get in the way of actual coding. It's not a drain on resources any more than DragonFlyBSD is a drain on the development of Inferno. Totally different space, requiring totally different mindsets and totally different skills.
I'd refine that to "they're not hard sciences - yet". There's a dispute that's older than me (by a long way) as to whether soft/social sciences and social sciences are sciences at all and I won't get into that. What I will say is that there is absolutely nothing in any physical science which strictly prohibits any of the soft/social sciences becoming hard sciences eventually. That anyone knows with any certainty.
I will add this proviso: "The Emperor's New Mind" by Roger Penrose would seem to imply that psychology can never be a hard science, since it does claim that the brain is a quantum computer which is irreducible to a deterministic model. That's not quite enough, since QM is perfectly good physics and yet not reducible to a deterministic model, but the brain is hellishly complex and if we can't model even trivial macroscale systems using QM we certainly won't be able to model something as convoluted as the human brain. We might not need to model it to quite that degree to be able to derive laws that are as good in the social disciplines as Newtonian mechanics is in the physical sciences, which would probably be good enough to qualify as a hard science, but we'd not be able to go beyond that point if Penrose is right.
However, as things stand, you are absolutely correct. The soft sciences (whether or not they really are science) often do not use the scientific method and frequently are more opinion-based than anything. In short, not merely pre-modern-science but pre-Socratic. It shouldn't take more than 2,500 years for them to catch up, though. Less, if they put in the fundamental research necessary.
For your last point, I'll offer a thought experiment I devised a while back. It assumes the existence of wormholes, which may or may not invalidate it. The idea is this. You start with the basics of Schrodinger's Cat experiment (a radioactive particle in a box) but place the detector (or cat and poison vial, if you like) in a different box. Link the two boxes together with a wormhole. Now, place the two boxes in different timeframes by accelerating one of them.
Let us say that the box that contains the detector is in the past, relative to the box containing the radioactive particle. The particle may or may not decay. In the many-worlds interpretation of quantum mechanics, both events will happen. Since the detector is in the past of this event, it cannot see which of the two events has happened. It would seem to be a bit like having a Y connector in a water pipe or an OR gate in logic - if either input has something going through it, there will be an output. This means that the detector will register an event, EVEN IF the observer who eventually opens the box does so in the universe in which the particle has not decayed.
This produces two possible options. Either the collapse of the waveform happens at the detector (in which case the particle's future is pre-determined - it HAS to decay or not decay according to what the detector says) OR the many-worlds interpretation is incorrect. Any other possibility requires the possibility of having an event in one universe caused by an event in a different universe. This would not violate causality overall, merely within the confines of that universe.
This thought experiment was something I put on the shelf for a while, but I got much more interested in it with Hawking's theory regarding information near black holes. The problem is that information cannot be created or destroyed in the standard model, not without some serious consequences (such as the total unraveling of causality). Hawking's argument is that in the many-worlds theory you can sum information across alternate universes, which means you get zero loss, which means there's no messy causality problems. But if you CAN sum information across universes then, provided wormholes exist, you can indeed Y-pipe those universes together.
Now, let's try extrapolating from this thought experiment. Let us say that the solution to the first problem is that the collapse of the waveforms does indeed migrate back in time to the detector. Let us also say that there are wormholes naturally occurring in the quantum foam, where the ends (since all particles are created in pairs in the foam and a wormhole is its own pair) are moving at some non-zero velocity with respect to each other. This would imply that a certain amount of information dribbles from one timeframe to another all the time, which in turn means that some non-zero amount of information would have to have dribbled back to T=0. The implication of this is that certain events in time are (as Doctor Who has claimed) indeed fixed. The split into the different many worlds has already happened, the waveform has already collapsed, what happens in this universe - with respect to those specific events - cannot change. All universes that fork from this one will have those events in common.
If all information dribbled back, then all possible outcomes would still happen but in any given universe they already have. The sum total would be totally non-deterministic even though any individual universe would be totally deterministic. Whilst solving any time-travel paradox and without technically violating the many-worlds concept, I'm not sure you'd get a lot of physicists to buy into a strict T=0 split. Although I would add that I'm not the first to propose that as a way of resolving QM issues. It has been proposed by better-qualified minds before, albeit for different reasons.
That is why courts sometimes (but not always) considers the mental age of the person as well as the physical age. It might be better if there were rules mandating that mental age always be considered in all legal matters (regardless of what the legal matter is) and it might be better if psychologists and neurologists could produce a sufficiently clear description of how to identify said mental age so that a judge or jury could actually evaluate any such claim. Then there's the question of whether a person's emotional age is necessarily the same as their mental age and how significant that would be if it is different.
In short, the situation is certainly a mess and the science is nowhere near evolved enough to offer any useful solution. I don't see the legal situation even being fixable until the science is fixed, simply because there has to be a foundation built on something and you've got to have some measure of confidence in that something - or at least more confidence than in what exists at the moment.
Politically, there's never going to be any interest in investing in the research needed. Way too great of a risk of a backlash of being "soft on crime", as there's bound to be more adults before the courts with a retarded mental age than there are children before the courts with an advanced mental age. Psychologists will never take the risk of doing the research on their own, as politics drives the professional boards. No businessman is likely to take a gamble on investing in something that will have no payback for them and puts them at serious risk of scandal. And nobody else has the resources or the qualifications. Which means you won't see a change until society itself wants to change. Historically, that's usually a hundred or so years after the problem is first noted, so 2110 is really the earliest any reform is likely to start.
Well, yes, that would be one possibility. (An outed informant could be followed, so that the person/patrol they communicate with can be ambushed, or the informant could be supplied with bad information - a tactic the allies used on numerous occasions in World War 2.) There may be others. Regardless of the specifics of what information has been exposed that could be exploited, if there is exploitable information then there will be a tactical "need to know" that differs from the strategic "need to know". That would put troops in danger in a way that the leaks themselves did not. It is extremely doubtful that the military could possibly process this discrepancy or find solutions that were acceptable in any sensible timeframe - it's too big, too bureaucratic and in many ways too naive. That means the only viable solution it has is to sacrifice some pawns on the battlefield, as creating a whole new concept of "need-to-know" is going to take a damn sight longer than finding out who actually needs to know.
Hmmm. I'm not certain of the precise rules, but what happens if someone is given authorization to access those documents from an internal source? Surely the control is on the document itself and not a specific instance of the document (since by accessing a document across a secure network you are creating a copy and it is not that specific copy you were given the access to). The rules, at least as outlined in the summary, would prohibit the access of the external Wikileaks copy regardless of any level of authorization they may have regarding any other copy.
In a way, that makes sense as the external copy is not authoritative and also if only authorized people accessed the file from there it could expose who had such authorization. (First rule of security is that you want minimal-to-no context to be unsecure.) On the other hand, the absolute logic of a blanket ban seems suspect to me. Also, there's another factor to consider. Whilst I sincerely doubt information as old as that leaked could have any tactical value in the present day, the military have said otherwise. (It tells me they're ignoring traditional military philosophy*, but there you go.) If their claim is correct, then that must surely create a problem for officers on the ground. The claim by the military implies that there are tactical or strategic decisions which rely on security through obscurity in order to work that have been exposed, but that those who may be placed in unnecessary danger as a result aren't going to know which decisions those are.
*"Art of War" and "Book of Five Rings" are standard military texts. Five Rings certainly states that you shouldn't repeat things to the point of being predictable and I'm pretty certain Art of War wasn't keen on over-reliance on too narrow a set of tactics. If documents from six or seven years ago can be used to accurately predict in advance the precise movements of troops today, I'd say there's a serious predictability problem.
Slashdot Mirror
Much more interesting (to me) is that the aircraft used (a DeHavilland DHC-3T) is a flying boat built in 1953. Looks great for fishing trips, but they were taking it through dangerous mountains - known for doing unmentionable things to aircraft - during a severe storm. Hands up all those who would want to be in the aircraft shown in the posting under those kinds of conditions. I feel certain that former Sen. Stevens has been in enough light aircraft (Alaska is big and the roads aren't) to know what you can and cannot do. What I cannot fathom is why he, with his knowledge and experience, would take that kind of a pointless, stupid risk.
As for O'Keefe, for all his time at NASA, I doubt he's enough time in aircraft of this vintage and size to know the risks. Being head honcho of an aerospace organization doesn't endow you with the kind of skill and knowledge needed. Nonetheless, he too should have been wary of flying in those conditions in an antique.
It is not an error. Repeat, it is not an error. This bill is indeed about censoring the obscene language in XXXX lager commercials.
Oh, much worse. But let's face it, the viewers are divided into two camps - those wanting spectacular disasters and those wanting spectacular and strategic clothing failures on female crew. Language is not a particular barrier to either of these...
Why do I seriously suspect that this will actually get written into a D&D module now...?!
You'll have to ask Arthur Scargill about the first of those, and the Forestry Commission for the latter. (The modern word "forest" actually started in England.) I'm fairly sure that when most of these numbers were drawn up, most UK mines were still open. That would likely mean the UK had as many - or more - coalminers than the US per capita. For lumberjacks, much the same argument applies. The amount of lumber produced in the UK at that time, per capita, would have been very comparable. (You must remember that the US figures are for the WHOLE of the US. Not many lumberjacks in Montana or North Dakota, but there are people. Well, in the case of North Dakota, person.)
I have no idea what +1, Interesting smallpox is. Presumably, it's a mutation of smallpox that has evolved the ability to post on Slashdot. In which case, we're safe. It can't spread. What's it going to reproduce with?
Aha! Found a buoy that is registering waves more frequently. Peaked at 7.5' a few hours back. Still relatively gentle swell, but not something for the faint-at-heart.
There's that and the difficulty in keeping a camera lens clean when the water spray freezes on impact. Current conditions (as of 10:47pm GMT) look pretty gentle - 3' waves are nothing. I'm fairly sure that rough seas would be nearer an average of 27-30'. For freak waves, 55.8' waves have been observed by the Netherlands lifeboat association. In the North Sea off Scotland, waves have been recorded as high as 61' for multiple and 95' for an individual wave. I don't know if the North Sea gets more of them than other regions, but Discovery might want to consider that NOBODY has ever done a "reality TV show" that included freak waves, and only one or two have ever been filmed at all. Ever. And by far cheaper equipment than Discovery can get hold of. As for the crew's extreme improbability of surviving, it's television! Besides, the ratings would surge like anything, as all the freaks out there wait for the next true-life disaster.
The banks can't afford Second Life accounts after giving so much in bonuses to top management.
In Netrek, the term is Ogging, after the greatest maniac ever to have played the game. (No, not me. I'm not even close.) There, they'll Ogg anything - space stations especially.
Don't knock those Second Life home owners. So long as their real-estate still has that value in the real world, those owners have the only property in the world the banks can't touch.
I wonder if the fan site has asked the marketing people to intervene. Usually, although lawyers can get lots of money, marketing people HAVE lots of money and that can make a difference to the upper echelons.
Alaskan crab fishing is ok as a "dangerous reality TV" show. Apparently the stats for Alaskan crab fishing is 356/100,000/year. That's a lot - getting on for 1 in every 100,000 per day. (US National Average workforce fatality rate is 7.0/100,000/year.) I wondered if North Sea fishing was worse - it has a vicious reputation and the North Sea has no landmass between it and the north pole. However, statistics indicates that the mortality rate is 151 x national average in the UK, and the UK's national average is 0.5 deaths per 100,000 people. That puts the North Sea fishermen at a paltry 76/100,000/year. Not safe, by any standards, but many times safer in absolute terms. In relative terms, the US' workforce fatality rate is 14x worse than that in the UK, but the Alaskan crab fishing is only 4.7x as deadly as North Sea fishing. By this standard, North Sea fishing is the deadliest fishing occupation relative to the health and safety of the country involved.
A valid point, but making sure things will work in the US should never dissuade anyone from looking at what works elsewhere -- especially if that includes the important elements of studying WHY it works there and what flaws were discovered when the method was implemented. ("Why" has to include not just the day-to-day why but also the pre-conditions that made it workable to begin with.) If all this is known and compatible pre-conditions exist in the US, then it should be possible to make the tweaks needed to adjust for any differences and to prevent the flaws discovered elsewhere.
There is nothing to stop one country learning from another. Learning is good. What is bad is when one country assumes either that what they have learned MUST apply - undiluted and unmodified - or must NOT apply.
Two or three. Since the post can't drop below -1, you only need to bring it up to +1 or +2 to be visible at most levels of filtering. On EARLIER versions of Slashdot, you could indeed bury a post permanently by taking it to -2. You had one person mod up by 1, the rest of your clique then modded down to get the post to -1, then the person who modded up posted, eliminating the +1. This took the post to -2, which is never visible and can therefore never be modded back up (except by sysops). This bug was fixed some time back.
A related quirk was that you could also get to the dizzying heights of +6 by the same method. The total number of +6-modded posts was extremely small, but they did exist. They also caught attention merely because every poster on the site knew damn well that should be impossible.
I agree, but formal education and the Internet (or television) are not mutually exclusive. In the UK, Open University is an extremely old (for mass communication, at least) and highly respected approach, albeit not as respected as the higher-end bricks-and-mortar facilities. Regardless of the fact that the lectures have been distributed for free for, what, 40? 50? years, there has been no evidence in the UK of the kind of shift Bill Gates has suggested. It would seem to follow that merely being free and widely circulated is insufficient. Further, UK citizens have, in general, had far more acceptance for this kind of system than the US - which is why PBS is in such poor shape and educational channels in general in the US are a bit of a disaster. (If Discovery can only afford to show 6 programs a quarter, at their size, and record even less than that, when the gameshow channels can afford to splurge on prizes bigger than the cost of recording a documentary every hour, you can tell something about what the priorities are.)
The answer is maybe. My thought experiment would suggest it may be possible for one possible outcome to alter the state of the universe in which a different outcome happened, which is similar to your own idea. If that is true, then the answer is necessarily yes. There may be other circumstances in which it could happen, but once you allow - even under the most restrictive of circumstances - a Y-piece to be added to the possible outcomes, then it becomes possible for a particle to go through one slit but register as though it went through the other.
To complicate things further, you can do diffraction over time as well as space. I found the original paper and two followups but Physical Review is subscription only. If it is possible to diffract in time as well as space, then it automatically follows that events don't just happen and then go away. The different points in time have to be interacting in some way. I do not fully understand the implications of that, but I would interpret it as meaning that it does indeed mean that it is possible to snake between the different possible worlds, or at least some subset of them.
If it was plain, then it would be more obvious to the rest of us. Since not a single one of the replies to me agreed on the issue - with me or each other - it is clear it is not plain. I will certainly look the guy up, but I sincerely doubt it will change my mind on whether the social sciences are currently or could potentially be a science.
Personally I agree with you, and if it's now pretty much solidly known then that's in line with what I'd have expected. My use of Penrose's claim is because I'm extremely wary of asserting that I'm right and a highly respected professor is wrong. (Wary, but I sometimes do that anyway, and I'm right often enough to make me wonder why they're the one with the cool title and tenure.)
I am extremely glad for your point, though, as it does help clarify things.
First, open source code is - and will be - used to kill people, whether you like it or not. If you don't contribute to such projects for moral reasons, you're not really helping those who use the software for moral purposes and you're not really hurting those who use it for immoral purposes.
Second, this is really a question of the level of indirection. If you contribute a patch to a kernel module which can be used by Linux and that version of Linux is downloaded and used by a third-part vendor which supplies the package to the DoD which, in turn, supplies it to the front-line, you've five levels of indirection. If we assume that each level merely doubled the number of people involved, then you would be 1/(2^5)th responsible for the patch being where it was. That would not make you 1/(2^5)th responsible on how it was used, though. You should really weight for that, putting it something closer to 1/(2^(2^5))th. In practice, given the size of the military, the size of the Linux userbase and the size of the Kernel development team, the dilution is closer to 128 rather than 2. So you're looking at 1/(2^(128^5))th of a share in the responsibility.
True, it is not zero, and I do have a serious issue with those who assume that dilution/indirection == zero responsibility, but at that kind of level it's so damn close to zero that it makes no serious odds. If you were to supply that same patch directly TO the front line, specifically knowing its intended use (or specifically not asking), that would be a different kettle of fish. There would be no dilution and no indirection.
Now, working with a project that the CIA probably will never use (they're a political organization, you think they'd use something if it would look like backing down?) and the military in general probably won't use (partly the Not Invented Here, partly because this is ultimately GOTS and the military are strictly prohibited from using GOTS over COTS, hence the absurd number of contractors and the absurd contract regulations and licensing issues) - meh. There's likely far more risk of death and destruction coming from using one of the NASA CFD packages and handing back a patch (as that might well be used for designing military aircraft).
This piece of software might actually be quite handy for the open source community, though. And the encryption community. There's lots of flamewars over stupid issues. The current one over on the SHA-3 mailing list is over whether NIST's security requirements should be met or if designers can "cheat" if nobody is likely to break the algorithm anyway and to be "secure" means to be slow. To me, this is stupid. You use algorithms that are "good enough" where you need the speed and you use algorithms that are "secure" when you need the security. Try getting that through some incredibly thick craniums some time. However, if a piece of software produced exactly the same advice, I can bet you anything that those same people would pay attention.
The same goes for whether Linux audio should be robust or realtime. GUIs should be KDE or Gnome. SELinux or GRSecurity. Logging FS versus Journaling. We'd be miles further along if the petty bullshit of one-upmanship didn't get in the way of actual coding. It's not a drain on resources any more than DragonFlyBSD is a drain on the development of Inferno. Totally different space, requiring totally different mindsets and totally different skills.
I'd refine that to "they're not hard sciences - yet". There's a dispute that's older than me (by a long way) as to whether soft/social sciences and social sciences are sciences at all and I won't get into that. What I will say is that there is absolutely nothing in any physical science which strictly prohibits any of the soft/social sciences becoming hard sciences eventually. That anyone knows with any certainty.
I will add this proviso: "The Emperor's New Mind" by Roger Penrose would seem to imply that psychology can never be a hard science, since it does claim that the brain is a quantum computer which is irreducible to a deterministic model. That's not quite enough, since QM is perfectly good physics and yet not reducible to a deterministic model, but the brain is hellishly complex and if we can't model even trivial macroscale systems using QM we certainly won't be able to model something as convoluted as the human brain. We might not need to model it to quite that degree to be able to derive laws that are as good in the social disciplines as Newtonian mechanics is in the physical sciences, which would probably be good enough to qualify as a hard science, but we'd not be able to go beyond that point if Penrose is right.
However, as things stand, you are absolutely correct. The soft sciences (whether or not they really are science) often do not use the scientific method and frequently are more opinion-based than anything. In short, not merely pre-modern-science but pre-Socratic. It shouldn't take more than 2,500 years for them to catch up, though. Less, if they put in the fundamental research necessary.
For your last point, I'll offer a thought experiment I devised a while back. It assumes the existence of wormholes, which may or may not invalidate it. The idea is this. You start with the basics of Schrodinger's Cat experiment (a radioactive particle in a box) but place the detector (or cat and poison vial, if you like) in a different box. Link the two boxes together with a wormhole. Now, place the two boxes in different timeframes by accelerating one of them.
Let us say that the box that contains the detector is in the past, relative to the box containing the radioactive particle. The particle may or may not decay. In the many-worlds interpretation of quantum mechanics, both events will happen. Since the detector is in the past of this event, it cannot see which of the two events has happened. It would seem to be a bit like having a Y connector in a water pipe or an OR gate in logic - if either input has something going through it, there will be an output. This means that the detector will register an event, EVEN IF the observer who eventually opens the box does so in the universe in which the particle has not decayed.
This produces two possible options. Either the collapse of the waveform happens at the detector (in which case the particle's future is pre-determined - it HAS to decay or not decay according to what the detector says) OR the many-worlds interpretation is incorrect. Any other possibility requires the possibility of having an event in one universe caused by an event in a different universe. This would not violate causality overall, merely within the confines of that universe.
This thought experiment was something I put on the shelf for a while, but I got much more interested in it with Hawking's theory regarding information near black holes. The problem is that information cannot be created or destroyed in the standard model, not without some serious consequences (such as the total unraveling of causality). Hawking's argument is that in the many-worlds theory you can sum information across alternate universes, which means you get zero loss, which means there's no messy causality problems. But if you CAN sum information across universes then, provided wormholes exist, you can indeed Y-pipe those universes together.
Now, let's try extrapolating from this thought experiment. Let us say that the solution to the first problem is that the collapse of the waveforms does indeed migrate back in time to the detector. Let us also say that there are wormholes naturally occurring in the quantum foam, where the ends (since all particles are created in pairs in the foam and a wormhole is its own pair) are moving at some non-zero velocity with respect to each other. This would imply that a certain amount of information dribbles from one timeframe to another all the time, which in turn means that some non-zero amount of information would have to have dribbled back to T=0. The implication of this is that certain events in time are (as Doctor Who has claimed) indeed fixed. The split into the different many worlds has already happened, the waveform has already collapsed, what happens in this universe - with respect to those specific events - cannot change. All universes that fork from this one will have those events in common.
If all information dribbled back, then all possible outcomes would still happen but in any given universe they already have. The sum total would be totally non-deterministic even though any individual universe would be totally deterministic. Whilst solving any time-travel paradox and without technically violating the many-worlds concept, I'm not sure you'd get a lot of physicists to buy into a strict T=0 split. Although I would add that I'm not the first to propose that as a way of resolving QM issues. It has been proposed by better-qualified minds before, albeit for different reasons.
That is why courts sometimes (but not always) considers the mental age of the person as well as the physical age. It might be better if there were rules mandating that mental age always be considered in all legal matters (regardless of what the legal matter is) and it might be better if psychologists and neurologists could produce a sufficiently clear description of how to identify said mental age so that a judge or jury could actually evaluate any such claim. Then there's the question of whether a person's emotional age is necessarily the same as their mental age and how significant that would be if it is different.
In short, the situation is certainly a mess and the science is nowhere near evolved enough to offer any useful solution. I don't see the legal situation even being fixable until the science is fixed, simply because there has to be a foundation built on something and you've got to have some measure of confidence in that something - or at least more confidence than in what exists at the moment.
Politically, there's never going to be any interest in investing in the research needed. Way too great of a risk of a backlash of being "soft on crime", as there's bound to be more adults before the courts with a retarded mental age than there are children before the courts with an advanced mental age. Psychologists will never take the risk of doing the research on their own, as politics drives the professional boards. No businessman is likely to take a gamble on investing in something that will have no payback for them and puts them at serious risk of scandal. And nobody else has the resources or the qualifications. Which means you won't see a change until society itself wants to change. Historically, that's usually a hundred or so years after the problem is first noted, so 2110 is really the earliest any reform is likely to start.
Well, yes, that would be one possibility. (An outed informant could be followed, so that the person/patrol they communicate with can be ambushed, or the informant could be supplied with bad information - a tactic the allies used on numerous occasions in World War 2.) There may be others. Regardless of the specifics of what information has been exposed that could be exploited, if there is exploitable information then there will be a tactical "need to know" that differs from the strategic "need to know". That would put troops in danger in a way that the leaks themselves did not. It is extremely doubtful that the military could possibly process this discrepancy or find solutions that were acceptable in any sensible timeframe - it's too big, too bureaucratic and in many ways too naive. That means the only viable solution it has is to sacrifice some pawns on the battlefield, as creating a whole new concept of "need-to-know" is going to take a damn sight longer than finding out who actually needs to know.
Hmmm. I'm not certain of the precise rules, but what happens if someone is given authorization to access those documents from an internal source? Surely the control is on the document itself and not a specific instance of the document (since by accessing a document across a secure network you are creating a copy and it is not that specific copy you were given the access to). The rules, at least as outlined in the summary, would prohibit the access of the external Wikileaks copy regardless of any level of authorization they may have regarding any other copy.
In a way, that makes sense as the external copy is not authoritative and also if only authorized people accessed the file from there it could expose who had such authorization. (First rule of security is that you want minimal-to-no context to be unsecure.) On the other hand, the absolute logic of a blanket ban seems suspect to me. Also, there's another factor to consider. Whilst I sincerely doubt information as old as that leaked could have any tactical value in the present day, the military have said otherwise. (It tells me they're ignoring traditional military philosophy*, but there you go.) If their claim is correct, then that must surely create a problem for officers on the ground. The claim by the military implies that there are tactical or strategic decisions which rely on security through obscurity in order to work that have been exposed, but that those who may be placed in unnecessary danger as a result aren't going to know which decisions those are.
*"Art of War" and "Book of Five Rings" are standard military texts. Five Rings certainly states that you shouldn't repeat things to the point of being predictable and I'm pretty certain Art of War wasn't keen on over-reliance on too narrow a set of tactics. If documents from six or seven years ago can be used to accurately predict in advance the precise movements of troops today, I'd say there's a serious predictability problem.
Criminal psychologst calls CEOs psychopaths
Sociologist/Criminologist calls CEOs sociopaths
Take your pick. Or maybe they're both. It would explain a lot.