The evidence for something-caused global warming is pretty strong. The evidence for humans as the cause of it? There isn't really any. It's not the sort of thing that offers much in the way of evidence - what on earth would such evidence look like? A giant signpost in the sky saying "Hey morons, you did this"?
We know the planet's getting a little bit hotter. We know what we're doing. The only thing we have connecting the two, so far, is a pile of theories, most of which disagree with each other. This isn't something you can set up a lab experiment to prove - the best we have are computer models which start out by saying "IF MY THEORY IS CORRECT... this is what's going on".
Many scientists in the field think that it's likely that human actions are responsible, which is why they're working on the subject (very few people ever set out to try and prove something which they did not originally think was true). That's a long, long way from having any evidence to back it up. Almost all of them will tell you this, if you bother to ask them (which the media usually doesn't).
We desperately need a better understanding of climate science - and I mean real understanding of what actually happens, rather than theories about what might be happening.
(What should we do about all this? Damned if I know. But it certainly won't help to lose perspective on just how little we really know about what's going on here.)
If you had read the comments on the story you link to, you would have learned that no current "global warming" theories are able to explain why that happened. Even the worst of the global-warming-doomsday predictions say that right now, today, the sea levels should have risen by an inch or two. Certainly not enough to swamp any islands worthy of the name.
Something else is going on there. Probably some kind of plate tectonic activity. There are numerous ex-inhabited islands that have sunk due to plate movements. It's a normal geological process.
A projectile will travel a ballistic path -- an ellipse.
This is true only as an approximation at low altitude. A projectile fired along a "ballistic" path that takes it to a sufficient height will have a reduced downwards acceleration due to gravity - it's only 9.8m/s^2 near the surface of the planet, that value goes down as you get higher. If it travels high/fast enough, this will cause it to deviate from its "natural" ellipse and follow an orbital path instead.
When does this happen? The ability to orbit is determined by the energy of the object, and only by the energy of the object (with the obvious exception that if it's aimed below the horizon, it will merely make a big hole in the ground). That's basic orbital mechanics - an object with kinetic and gravitional potential energy that is equal to the required energy level for this planet is defined as being in a "closed orbit", and will go around the planet until that state changes. (An object with more energy than this is defined as being in an "open orbit", and will continue to move away from the planet)
This is basic Newtonian physics - if you set up the ballistic equations using the true function for gravity (G*m1*m2/r), instead of the constant approximation (9.8), you get the behaviour described above.
If any government-funded entity did this, the patent-funded corporations would scream "unfair competition!" and send their hordes of patent-funded lawyers and lobbyists to get them shut down.
Is that just an omission on the part of the reporter or do you really believe you have no moral responsibility to intervene when you see someone doing something wrong?
A "moral responsibility to intervene when you see someone doing something wrong" is the only reason which has ever been used to censor things. The concept that it is moral to intervene when somebody else breaches your concept of 'right' and 'wrong' is the very antithesis of liberalism.
With the new developments vastly increasing the ease of reproduction of buildings, and the sudden upsurge in building piracy costing the industry over $10bn per year, it is necessary to implement strong rights management in order to prevent people from illegally producing buildings without paying a license fee to the architectural design firm. To provide fair compensation to the children of architects, new laws are being introduced that require all buildings to be made from approved construction materials that implement the StaysUpForSure protocol, which allows software monitoring and control of every component, in the "Fair House Prices for Children Act".
The "Walls" house operating software (included with every new house purchase) scans all components of the house, several times a second, to check for unauthorised modifications or attempted duplication. It contacts the central licensing servers once a day to ensure that this design of house is licensed for construction at this location, validated against its built-in GPS receiver. If the GPS receiver cannot receive a signal, or the licensing server does not report that the building is approved at the current location, or the component validator detects unauthorised modifications, then the software will signal all the construction materials to shut down, causing the house to collapse and protecting you from the dangers of building piracy.
Building insurance companies welcomed the move, saying: "Before now, when a house fell down, we had to spend money on careful investigations to identify whether the house was constructed from properly licensed blueprints - but now we can be sure that any collapsed house is the result of building piracy, which voids the insurance policy".
The public interest is in having this matter resolved in a reasonable time frame.
That's not the primary reason why this attempt is likely to fail. Judges have basically four priorities when deciding what they think about something, in order of decreasing priority:
Lawyers screwing with them
Their caseload and schedule
Justice and the law
Public interest
Delaying tactics fall squarely under #2. Judges loathe delay. A delayed case is a case that they're going to have to keep working on, and cases always get more complicated the longer they last. (Judges typically don't mind working hard, but they absolutely hate having to work on the same damn thing every week for several years)
It's now nearly 4 years later. The two judges in this case know why it's been dragging on this long. They will be *severely* pissed off about that, and not at all amused at proposals (from either party) to make it drag out any longer.
Breeder reactors, however, generate weapons grade plutonium, which is why they are considered more of a proliferation risk.
All of this is assuming that you want to make a nuclear detonation. They're technologically pretty, but a stupid method for terrorists. A far more sensible approach would be to take your regular old reactor-grade uranium or plutonium, strap it to a large mass (say the contents of three or four suitcases) of conventional explosives, and blow it up. This is known as a "dirty bomb" - the yield is nowhere near that of a nuclear bomb, but it still sprays lethal radioactive material over a wide area. Three or four of these in a large city would kill a large portion of the population and render the entire city uninhabitable for decades. No specialist knowledge or enrichment is necessary - it's just a box of uranium on top of a big explosion. And after they went off, nobody would mention 9/11 again except in history classes.
Yes, the concern is completely ridiculous. Plutonium is not the problem. The tons of uranium that go missing every year? That's the problem. We don't know who has it. They can't be planning to do anything good with it (you can't hide a nuclear power plant).
Yes, but the absolutely daft US regulations forbid extracting plutonium from spent fuel. After all, it might make it easier for terrists to get holda some and make a nukular bomb.
To be fair, this is not a modern policy. This was made policy by Jimmy Carter, and it was well acknowledged that he was doing it with a wink and nod to the anti-nuclear-energy lobby.
And it's widely known to be nonsense. If you want to make bombs, uranium is a perfectly adequate material - plutonium is not required. The US used plutonium during the Cold War because it produces more destruction per dollar - a pure cost decision, which terrorists aren't likely to care about. Of all the nuclear bombs ever deployed outside test sites (two), exactly half (one) were uranium bombs.
1) We can. It's just not necessarly economic to pull it out.
Why not?
Because mining more fresh uranium is cheaper.
Yeah, it's that fucked up. We aren't burying this stuff because we have to. We aren't doing it because continuing to use it as fuel wouldn't make money. We're doing it because burying the spent fuel and mining fresh fuel improves the bottom line of the power companies - the net cost is lower than reprocessing the spent fuel.
At some point in the future (unknown, depends how many more uranium deposits we find - but at current growth rates, the ones we know about won't last 100 years), this will change. And then we're going to be digging up all this "waste" that we're burying, because we want to use it as fuel.
It's my understanding that the rules guide the TSA staff, not the airlines. Remember, the airport checkpoints are no longer run by private companies, they're run by the TSA.
That wasn't true at the time Gilmore filed his suit (which is the time the courts are primarily interested in) - but anyway, it makes no difference. The TSA pursues legal action against its own staff who break its rules.
What grounds would Gilmore have to get a court to order the airline to hand over their trade secrets?
The grounds of that information being relevant to his case. That is a fundamental principle of the court system - the other party must hand over all relevant information. Trade secrets do not obstruct this at all - the judge will merely seal the relevant court papers, so that nobody other than the parties can see them (and the party receiving the information is obliged not to use it for anything outside the case).
There are no valid legal grounds on which anything can be withheld from the other party in a court case, if the party can show that the information is relevant to the case (you don't have to prove that, but you do have to demonstrate some kind of evidence that the information is relevant). The only exception is when the party who wants to keep something secret is the government.
As for the upgrade thing - don't we know it. Office 2007 rollout isn't going to happen before September, if not 2008 (getting the teachers to put time in learning the new interface so they can teach the kids is the hard part!).
Stop.
Let's examine this more closely.
Teachers have to put in time to learn the new interface, or they can't teach the pupils how to use it. Therefore, their knowledge of how previous versions of Office work is not adequate for them to do their jobs, and teaching the kids how to use the old version is not adequate for the kids to use the new version.
This year, you have been teaching the kids how to use some unspecified version of Office, which is older than 2007.
These kids will not see the workplace for several years yet. When they get there, Office 2007 (or later) is expected to be the 'norm'.
From the paragraph above, teaching the kids how to use older versions of Office does not adequately prepare them to use Office 2007.
WHAT POSSIBLE USE IS WHAT YOU WERE TEACHING THE KIDS THIS YEAR?
Seriously. If this argument is valid, then it is an admission that what you have been teaching the kids is useless to them in the future. The existence of this contradiction points to a deep and major need to re-examine the fundamental concepts of what the kids are supposed to be learning.
I've talked to the admin staff at three different UK universities who all had major Linux/UNIX deployments, at various times over the past 5 years.
All three had been approached by Microsoft, offering them special deals and what amounted to hard cash.
I fully expect that Microsoft has approached every single university in the UK who wasn't already all-Windows, with the same offers.
UK universities are having a hard time making their budgets balance these days, with reduced government funding and instructions to make things work on their own.
They obviously don't call it a law, because, well, it isn't a law (you can't be arrested or fined for disobeying it).
The secret law says what the airlines have to do. The airlines can and will be fined for disobeying it. You can't be arrested or fined under this law because it just doesn't apply to you directly - it's a set of rules about how the airline has to treat you. (I agree with your implication that a law is most reasonably defined as any rule made up by the government for which somebody can be fined or arrested - I just think this rule meets those requirements)
It's unclear to me what the difference is between the airlines making the rules and the TSA doing it.
If the airlines made the rules, then it wouldn't be a secret, and the court would just have ordered the airline to hand over their policies to Gilmore, so that he can see for himself if they're legal and argue his case in court. When the TSA makes the rules, this doesn't happen. That's the difference.
And when will we see a distro that incorporates bittorrent into its packaging download system?;-) (Slightly joking on that last one. I've no idea if it would be appropriate
It wouldn't. A protocol vaguely like bittorrent could work, but bittorrent itself is too focused on smaller, more centralised efforts. You have to realise that any large distribution cannot operate with anything less than several dozen mirrors, and really needs a figure approaching ~100 to maintain good performance, due to the sheer size of the total network load. While bittorrent can operate on those levels, the CPU demands on the tracker would be astronomical - too much for any reasonable server to handle (and while the trackerless variant of bittorrent works, it doesn't work *that* well). You could run multiple smaller trackers... but then you'd lose most of the benefit of bittorrent, and you'd have to come up with a way to distribute clients between them, which is generally seen as an impractical approach.
Nobody has yet invented a swarming system which can cope with the kind of traffic levels you see in these systems using the kind of resources that the large distributions have available. It will probably happen someday, but it hasn't happened yet, and it's not going to be the current version of bittorrent that does it. (If a later version of bittorrent were to implement an efficient distributed tracker system - and this is *hard* - that could possibly do it)
There is also the secondary problem that most of the distributions take advantage of the dozens of large free mirror sites around the world, and those sites only support http and ftp. A solution would have to be found. You can't just tell those people to install bittorrent, as they will ignore you.
You're stuck in the 1970s. Lambda calculus is an old formalism. It has some interesting applications, but it's still fundamentally old. We have better methods now. Operational semantics and its variations are the main ones here. They're far more powerful systems than the lambda calculus, and they work just fine for imperative languages, proving the same sort of things. You don't have to stick to any particular rules. All known features of modern real-world languages have been modelled sufficiently in at least one variation of operational semantics (the different variations deal with things like non-trivial type systems, which need special treatment - it's really tricky to prove things about a C++ template class, but it's doable if you use a sufficiently powerful formalism).
There are some interesting advantages to pure functional languages, but this is not one of them. It's just that back in the 1970s, the problem of how to do it for non-functional languages hadn't been solved yet. Nowadays, it has been solved - so you can stick with easy languages like C and INTERCAL, and don't have to deal with lisp variants.
You cannot write a program in a turing-complete language to determine if another program in a turing-complete language is 100% secure. It's trivially reducable to the halting problem. You can't automate it, at least not with 100% accuracy, in every case.
Congratulations, you have won today's "Ignorant undergraduate misunderstanding of the Halting problem" prize.
You're wrong on every significant point. You can write a program in a turing-complete language to determine if another program in a turing-complete language is 100% secure. The way that you specified implementing such a program is flawed. The Halting problem says that there exist certain ways to specify a problem which admit no solution, even though a solution to the problem exists. It does not say that there exist no alternative ways to specify the problem which do admit solutions. People always seem to think that the purpose of the Halting proof was to demonstrate that the real-world problem couldn't be solved - this is wrong. The purpose of the proof was merely to demonstrate that there exist certain non-trivial, interesting mathematical problem specifications which don't have solutions. This has interesting results in computability theory. It has very little relevance to the question of what sort of software we can write. It's all about how you reduce the real-world problem into a mathematical problem specification.
Assume you have an algorithm (however complex) that can determine if a program in some turing complete language is secure, call it IsSecure(). IsSecure() is provably secure, because you've ran it on itself.
Now, write a program that has a security hole if and only if IsSecure() returns true:
#Program A start(input) {
if(IsSecure(input))
ExposeSecurityHoleInSelf()
else
#The hole must be in the function IsSecure(), which is silly, because you've used IsSecure() to secure IsSecure() }
IsSecure returns false when passed this program as input. It doesn't matter that you think the answer is silly. This program is not secure because there exists a call to ExposeSecurityHoleInSelf in it and IsSecure failed to prove that this call was unreachable, or just didn't give a damn that it was unreachable. That is defined as an insecure program for the purpose of the IsSecure function. By specifying the problem in this way, we admit the possibility of a solution, and the Halting problem is avoided.
In most cases, the Halting problem can be avoided in this manner. Nothing compels you to define your program as having no false positives. For the purposes of automated security validation, false positives are not a serious problem - we can easily write the program in a manner that can be proven secure by a given prover. We don't have to accept arbitrary programs as input.
In practice, we don't do it like this. The function we use in the real world is is_proof_of_security_valid(), and it takes two inputs - a program and a proof of the program's security. The function checks that the proof is valid for this program. The proof itself is generated semi-automatically, but some parts are supplied by humans - typically via markup in the program's source (lint tags are a classic example of this sort of thing). It's much easier to write the thing this way.
The reason there is so much legacy code about is because that code has been around for some years, is proven and is bug free.
It's more like "because the people who would have to make the decision to replace it don't care about the bugs and see no point spending money on a replacement". They have low-paid grunts to take care of the bugs for them - and the labour costs are insignificant compared to the cost of the maintainance contract on that hardware.
You are failing to include the cost of the LTO-3 Tape DRIVE. Which, in itself, will cost around $3000 for a cheap one (thanks, CDW) and require a SCSI connection.
And a maintainance contract, because the damn things are fragile piles of barely contained self-destruction, and break at least once a year. And an intern to care for and feed the stupid thing with tapes, because you can't afford to buy a robot to do it, the sysadmins threatened to resign if you made them do it, and the users cannot understand instructions like "If you have deleted all your files, tell us BEFORE two weeks have elapsed so that we don't have to go and fetch a tape". Estimated annual cost, around $10k (assuming that you can get some work out of the intern the rest of the time, so only part of their salary is counted).
Tapes aren't something you want to buy and run yourself - they're something that you want to get somebody else to deal with for you. At this point, off-site data vault companies start to look better. (Send them your data via VPN or just courier a drive to them every week)
Datacenters don't necessarily want larger disks. Frequently, they are performance oriented and are more interested in spreading their dataset across a larger number of spindles for increased performance.
Higher data density == faster. There are real, serious physical limits in the speed at which a hard drive can spin and still be reliable. That's why your desktop drives are still spinning at the same speed they were 10 years ago - we can push them a little faster and still be reliable, but the cost goes up significantly, because the increased mechanical stress requires much tighter production tolerances. To get data through the head faster, since you cannot really alter the rate at which disk surface passes under a head, you have to pack more data into that surface.
There was a time when the answer was things like "use more heads". We did all those things. Now we must pack more data into the same physical area, if we want to continue improving performance at the rate we are used to.
Coincidentally, boosting the data density is how you increase the capacity of the disk.
So yes, datacenters want larger disks. Datacenter managers may not be aware of this, but it's true.
(At any given technology level, a manufacturer may have larger+slower models and smaller+faster models - that's irrelevant to the subject here. The process of research, which this article is about, will have to drive both performance and capacity up in tandem - it's not an either/or thing)
Compilers can also deal with optimization in RISC architectures more easily
This is a dead giveaway that the author is just stabbing at the wind. Scheduling is no more complex with CISC than with RISC. In fact, some compilation can be optimized even better by specialized CISC instructions that happen frequently. This is an ancient debate that is a tie.
No, it's a loss - for both sides (although the RISC crowd were a little more wrong). Theoretically the RISC crowd should have been right, as a RISC architecture lets the compiler do more optimisation ahead of time. In practice, nobody could figure out how to write such a compiler, which is why we don't use actual RISC chips any more. CISC is just what happens when you don't think about your ISA, so we don't use those any more either, except for x86.
All modern chip designs are hybrids, which do not fit into the historical RISC/CISC debate at all - they use the bits of RISC that we figured out how to compile for, and discard the rest. They typically include a set of specialised complex instructions for specific problems that the chip has to tackle (that may be aimed at performance or memory usage or power consumption or whatever). Compilers mix and match as best they can.
Modern CPUs have sufficiently complex decoder units at the front of sufficiently deep pipelines that the programmer-visible ISA has very little impact on the performance-critical aspects of the architecture. You need look no further than the vast architectural difference between AMD and Intel, or even Intel and Intel (P4 vs. Core) for the proof.
That doesn't prove any such thing - and in fact, the statement is wrong. The programmer-visible ISA does have a significant impact on the performance-critical aspects of the architecture. Not boring stuff, like SSE instructions, but important stuff like the guarantees about memory write ordering and things like that. An x86 chip must pretend that each instruction is fully completed before the next one executes, and has to jump through all kinds of complicated hoops to run them out-of-order internally. A POWER4 chip (now old, but it's one I know off the top of my head) typically executes instructions in groups of four or five, with little or no guarantees about the order in which the instructions within a group are executed - no hoop-jumping needed, the ISA presented lets the compiler handle all that. That makes a significant difference - and there's a whole bunch of similar historical rules that weigh down x86 but aren't present in PowerPC. You cannot decode away restrictions like this. Slapping a big pipeline on the front does not help - Intel thought it would, so they built the P4 around that principle, and then AMD gave them a resounding beating because Intel were wrong (possibly their worst design blunder ever). They have abandoned the overly-extended pipeline in the Core architecture for precisely this reason.
The real reason why this doesn't matter is far simpler. Intel and AMD put millions of x86 chips out each year. That lets them scale their production larger. Higher scale of production gives you lower costs - so even with the estimated 20%-30% (popular ballpark figure, nobody's entirely sure) performance loss from the crappy ISA, Intel and AMD get better performance per dollar than the architecturally superior competition. It's pure economics, nothing to do with the actual chips at all. Raw performance is not important; price/performance is what counts.
If anybody ever comes up with an ISA which gives you a fundamental advantage of a factor of 2 or more, this will change dramatically. So far, nobody has managed anything near that - but the research models are getting better all the time. Someday it might happen.
You have to realise that this is not a "computer" in the sense that you think of one.
This is an abnormally large cellphone. Its feature set is roughly equivalent to what you'd find on a modern 'smart' phone.
They don't market it like that because you'd have a hard time justifying giving free cellphones away to kids as a way of improving their lives. They sell it as a 'laptop', even though it really isn't what people think of when they use the term. It could never survive in a free market - this is a monopoly-based device. Like any device targeting a monopoly market, it is less than ideal.
Frankly, I think the OLPC project's vision is rather small. It could have been so much more.
But hey, you know what they say - give a man a fish, and he'll be fed today. Teach a man how to fish, and he'll sit in a boat and drink beer all day. For some reason, people think that is an appropriate objective, and aim no higher.
First, for corona discharge to occur at all requires thousands of volts of energy.
Buuu! Energy is not measured in volts. Attempts at debunking things while failing to make basic electrical sense are not impressive. Try harder next time.
Slashdot Mirror
The evidence for something-caused global warming is pretty strong. The evidence for humans as the cause of it? There isn't really any. It's not the sort of thing that offers much in the way of evidence - what on earth would such evidence look like? A giant signpost in the sky saying "Hey morons, you did this"?
We know the planet's getting a little bit hotter. We know what we're doing. The only thing we have connecting the two, so far, is a pile of theories, most of which disagree with each other. This isn't something you can set up a lab experiment to prove - the best we have are computer models which start out by saying "IF MY THEORY IS CORRECT... this is what's going on".
Many scientists in the field think that it's likely that human actions are responsible, which is why they're working on the subject (very few people ever set out to try and prove something which they did not originally think was true). That's a long, long way from having any evidence to back it up. Almost all of them will tell you this, if you bother to ask them (which the media usually doesn't).
We desperately need a better understanding of climate science - and I mean real understanding of what actually happens, rather than theories about what might be happening.
(What should we do about all this? Damned if I know. But it certainly won't help to lose perspective on just how little we really know about what's going on here.)
If you had read the comments on the story you link to, you would have learned that no current "global warming" theories are able to explain why that happened. Even the worst of the global-warming-doomsday predictions say that right now, today, the sea levels should have risen by an inch or two. Certainly not enough to swamp any islands worthy of the name.
Something else is going on there. Probably some kind of plate tectonic activity. There are numerous ex-inhabited islands that have sunk due to plate movements. It's a normal geological process.
This is true only as an approximation at low altitude. A projectile fired along a "ballistic" path that takes it to a sufficient height will have a reduced downwards acceleration due to gravity - it's only 9.8m/s^2 near the surface of the planet, that value goes down as you get higher. If it travels high/fast enough, this will cause it to deviate from its "natural" ellipse and follow an orbital path instead.
When does this happen? The ability to orbit is determined by the energy of the object, and only by the energy of the object (with the obvious exception that if it's aimed below the horizon, it will merely make a big hole in the ground). That's basic orbital mechanics - an object with kinetic and gravitional potential energy that is equal to the required energy level for this planet is defined as being in a "closed orbit", and will go around the planet until that state changes. (An object with more energy than this is defined as being in an "open orbit", and will continue to move away from the planet)
This is basic Newtonian physics - if you set up the ballistic equations using the true function for gravity (G*m1*m2/r), instead of the constant approximation (9.8), you get the behaviour described above.
If any government-funded entity did this, the patent-funded corporations would scream "unfair competition!" and send their hordes of patent-funded lawyers and lobbyists to get them shut down.
Big money defends itself.
A "moral responsibility to intervene when you see someone doing something wrong" is the only reason which has ever been used to censor things. The concept that it is moral to intervene when somebody else breaches your concept of 'right' and 'wrong' is the very antithesis of liberalism.
With the new developments vastly increasing the ease of reproduction of buildings, and the sudden upsurge in building piracy costing the industry over $10bn per year, it is necessary to implement strong rights management in order to prevent people from illegally producing buildings without paying a license fee to the architectural design firm. To provide fair compensation to the children of architects, new laws are being introduced that require all buildings to be made from approved construction materials that implement the StaysUpForSure protocol, which allows software monitoring and control of every component, in the "Fair House Prices for Children Act".
The "Walls" house operating software (included with every new house purchase) scans all components of the house, several times a second, to check for unauthorised modifications or attempted duplication. It contacts the central licensing servers once a day to ensure that this design of house is licensed for construction at this location, validated against its built-in GPS receiver. If the GPS receiver cannot receive a signal, or the licensing server does not report that the building is approved at the current location, or the component validator detects unauthorised modifications, then the software will signal all the construction materials to shut down, causing the house to collapse and protecting you from the dangers of building piracy.
Building insurance companies welcomed the move, saying: "Before now, when a house fell down, we had to spend money on careful investigations to identify whether the house was constructed from properly licensed blueprints - but now we can be sure that any collapsed house is the result of building piracy, which voids the insurance policy".
That's not the primary reason why this attempt is likely to fail. Judges have basically four priorities when deciding what they think about something, in order of decreasing priority:
Delaying tactics fall squarely under #2. Judges loathe delay. A delayed case is a case that they're going to have to keep working on, and cases always get more complicated the longer they last. (Judges typically don't mind working hard, but they absolutely hate having to work on the same damn thing every week for several years)
It's now nearly 4 years later. The two judges in this case know why it's been dragging on this long. They will be *severely* pissed off about that, and not at all amused at proposals (from either party) to make it drag out any longer.
All of this is assuming that you want to make a nuclear detonation. They're technologically pretty, but a stupid method for terrorists. A far more sensible approach would be to take your regular old reactor-grade uranium or plutonium, strap it to a large mass (say the contents of three or four suitcases) of conventional explosives, and blow it up. This is known as a "dirty bomb" - the yield is nowhere near that of a nuclear bomb, but it still sprays lethal radioactive material over a wide area. Three or four of these in a large city would kill a large portion of the population and render the entire city uninhabitable for decades. No specialist knowledge or enrichment is necessary - it's just a box of uranium on top of a big explosion. And after they went off, nobody would mention 9/11 again except in history classes.
Yes, the concern is completely ridiculous. Plutonium is not the problem. The tons of uranium that go missing every year? That's the problem. We don't know who has it. They can't be planning to do anything good with it (you can't hide a nuclear power plant).
And it's widely known to be nonsense. If you want to make bombs, uranium is a perfectly adequate material - plutonium is not required. The US used plutonium during the Cold War because it produces more destruction per dollar - a pure cost decision, which terrorists aren't likely to care about. Of all the nuclear bombs ever deployed outside test sites (two), exactly half (one) were uranium bombs.
Because mining more fresh uranium is cheaper.
Yeah, it's that fucked up. We aren't burying this stuff because we have to. We aren't doing it because continuing to use it as fuel wouldn't make money. We're doing it because burying the spent fuel and mining fresh fuel improves the bottom line of the power companies - the net cost is lower than reprocessing the spent fuel.
At some point in the future (unknown, depends how many more uranium deposits we find - but at current growth rates, the ones we know about won't last 100 years), this will change. And then we're going to be digging up all this "waste" that we're burying, because we want to use it as fuel.
That wasn't true at the time Gilmore filed his suit (which is the time the courts are primarily interested in) - but anyway, it makes no difference. The TSA pursues legal action against its own staff who break its rules.
The grounds of that information being relevant to his case. That is a fundamental principle of the court system - the other party must hand over all relevant information. Trade secrets do not obstruct this at all - the judge will merely seal the relevant court papers, so that nobody other than the parties can see them (and the party receiving the information is obliged not to use it for anything outside the case).
There are no valid legal grounds on which anything can be withheld from the other party in a court case, if the party can show that the information is relevant to the case (you don't have to prove that, but you do have to demonstrate some kind of evidence that the information is relevant). The only exception is when the party who wants to keep something secret is the government.
Stop.
Let's examine this more closely.
Teachers have to put in time to learn the new interface, or they can't teach the pupils how to use it. Therefore, their knowledge of how previous versions of Office work is not adequate for them to do their jobs, and teaching the kids how to use the old version is not adequate for the kids to use the new version.
This year, you have been teaching the kids how to use some unspecified version of Office, which is older than 2007.
These kids will not see the workplace for several years yet. When they get there, Office 2007 (or later) is expected to be the 'norm'.
From the paragraph above, teaching the kids how to use older versions of Office does not adequately prepare them to use Office 2007.
WHAT POSSIBLE USE IS WHAT YOU WERE TEACHING THE KIDS THIS YEAR?
Seriously. If this argument is valid, then it is an admission that what you have been teaching the kids is useless to them in the future. The existence of this contradiction points to a deep and major need to re-examine the fundamental concepts of what the kids are supposed to be learning.
I've talked to the admin staff at three different UK universities who all had major Linux/UNIX deployments, at various times over the past 5 years.
All three had been approached by Microsoft, offering them special deals and what amounted to hard cash.
I fully expect that Microsoft has approached every single university in the UK who wasn't already all-Windows, with the same offers.
UK universities are having a hard time making their budgets balance these days, with reduced government funding and instructions to make things work on their own.
It is obvious how this one works.
The secret law says what the airlines have to do. The airlines can and will be fined for disobeying it. You can't be arrested or fined under this law because it just doesn't apply to you directly - it's a set of rules about how the airline has to treat you. (I agree with your implication that a law is most reasonably defined as any rule made up by the government for which somebody can be fined or arrested - I just think this rule meets those requirements)
If the airlines made the rules, then it wouldn't be a secret, and the court would just have ordered the airline to hand over their policies to Gilmore, so that he can see for himself if they're legal and argue his case in court. When the TSA makes the rules, this doesn't happen. That's the difference.
The airlines and the TSA claim that it is a law written by the TSA. They also claim that the law is a secret and you are not allowed to see it.
Hence the lawsuit. Yes, secret laws make a difference.
It wouldn't. A protocol vaguely like bittorrent could work, but bittorrent itself is too focused on smaller, more centralised efforts. You have to realise that any large distribution cannot operate with anything less than several dozen mirrors, and really needs a figure approaching ~100 to maintain good performance, due to the sheer size of the total network load. While bittorrent can operate on those levels, the CPU demands on the tracker would be astronomical - too much for any reasonable server to handle (and while the trackerless variant of bittorrent works, it doesn't work *that* well). You could run multiple smaller trackers... but then you'd lose most of the benefit of bittorrent, and you'd have to come up with a way to distribute clients between them, which is generally seen as an impractical approach.
Nobody has yet invented a swarming system which can cope with the kind of traffic levels you see in these systems using the kind of resources that the large distributions have available. It will probably happen someday, but it hasn't happened yet, and it's not going to be the current version of bittorrent that does it. (If a later version of bittorrent were to implement an efficient distributed tracker system - and this is *hard* - that could possibly do it)
There is also the secondary problem that most of the distributions take advantage of the dozens of large free mirror sites around the world, and those sites only support http and ftp. A solution would have to be found. You can't just tell those people to install bittorrent, as they will ignore you.
You're stuck in the 1970s. Lambda calculus is an old formalism. It has some interesting applications, but it's still fundamentally old. We have better methods now. Operational semantics and its variations are the main ones here. They're far more powerful systems than the lambda calculus, and they work just fine for imperative languages, proving the same sort of things. You don't have to stick to any particular rules. All known features of modern real-world languages have been modelled sufficiently in at least one variation of operational semantics (the different variations deal with things like non-trivial type systems, which need special treatment - it's really tricky to prove things about a C++ template class, but it's doable if you use a sufficiently powerful formalism).
There are some interesting advantages to pure functional languages, but this is not one of them. It's just that back in the 1970s, the problem of how to do it for non-functional languages hadn't been solved yet. Nowadays, it has been solved - so you can stick with easy languages like C and INTERCAL, and don't have to deal with lisp variants.
Congratulations, you have won today's "Ignorant undergraduate misunderstanding of the Halting problem" prize.
You're wrong on every significant point. You can write a program in a turing-complete language to determine if another program in a turing-complete language is 100% secure. The way that you specified implementing such a program is flawed. The Halting problem says that there exist certain ways to specify a problem which admit no solution, even though a solution to the problem exists. It does not say that there exist no alternative ways to specify the problem which do admit solutions. People always seem to think that the purpose of the Halting proof was to demonstrate that the real-world problem couldn't be solved - this is wrong. The purpose of the proof was merely to demonstrate that there exist certain non-trivial, interesting mathematical problem specifications which don't have solutions. This has interesting results in computability theory. It has very little relevance to the question of what sort of software we can write. It's all about how you reduce the real-world problem into a mathematical problem specification.
IsSecure returns false when passed this program as input. It doesn't matter that you think the answer is silly. This program is not secure because there exists a call to ExposeSecurityHoleInSelf in it and IsSecure failed to prove that this call was unreachable, or just didn't give a damn that it was unreachable. That is defined as an insecure program for the purpose of the IsSecure function. By specifying the problem in this way, we admit the possibility of a solution, and the Halting problem is avoided.
In most cases, the Halting problem can be avoided in this manner. Nothing compels you to define your program as having no false positives.
For the purposes of automated security validation, false positives are not a serious problem - we can easily write the program in a manner that can be proven secure by a given prover. We don't have to accept arbitrary programs as input.
In practice, we don't do it like this. The function we use in the real world is is_proof_of_security_valid(), and it takes two inputs - a program and a proof of the program's security. The function checks that the proof is valid for this program. The proof itself is generated semi-automatically, but some parts are supplied by humans - typically via markup in the program's source (lint tags are a classic example of this sort of thing). It's much easier to write the thing this way.
It's more like "because the people who would have to make the decision to replace it don't care about the bugs and see no point spending money on a replacement". They have low-paid grunts to take care of the bugs for them - and the labour costs are insignificant compared to the cost of the maintainance contract on that hardware.
And a maintainance contract, because the damn things are fragile piles of barely contained self-destruction, and break at least once a year. And an intern to care for and feed the stupid thing with tapes, because you can't afford to buy a robot to do it, the sysadmins threatened to resign if you made them do it, and the users cannot understand instructions like "If you have deleted all your files, tell us BEFORE two weeks have elapsed so that we don't have to go and fetch a tape". Estimated annual cost, around $10k (assuming that you can get some work out of the intern the rest of the time, so only part of their salary is counted).
Tapes aren't something you want to buy and run yourself - they're something that you want to get somebody else to deal with for you. At this point, off-site data vault companies start to look better. (Send them your data via VPN or just courier a drive to them every week)
Higher data density == faster. There are real, serious physical limits in the speed at which a hard drive can spin and still be reliable. That's why your desktop drives are still spinning at the same speed they were 10 years ago - we can push them a little faster and still be reliable, but the cost goes up significantly, because the increased mechanical stress requires much tighter production tolerances. To get data through the head faster, since you cannot really alter the rate at which disk surface passes under a head, you have to pack more data into that surface.
There was a time when the answer was things like "use more heads". We did all those things. Now we must pack more data into the same physical area, if we want to continue improving performance at the rate we are used to.
Coincidentally, boosting the data density is how you increase the capacity of the disk.
So yes, datacenters want larger disks. Datacenter managers may not be aware of this, but it's true.
(At any given technology level, a manufacturer may have larger+slower models and smaller+faster models - that's irrelevant to the subject here. The process of research, which this article is about, will have to drive both performance and capacity up in tandem - it's not an either/or thing)
No, it's a loss - for both sides (although the RISC crowd were a little more wrong). Theoretically the RISC crowd should have been right, as a RISC architecture lets the compiler do more optimisation ahead of time. In practice, nobody could figure out how to write such a compiler, which is why we don't use actual RISC chips any more. CISC is just what happens when you don't think about your ISA, so we don't use those any more either, except for x86.
All modern chip designs are hybrids, which do not fit into the historical RISC/CISC debate at all - they use the bits of RISC that we figured out how to compile for, and discard the rest. They typically include a set of specialised complex instructions for specific problems that the chip has to tackle (that may be aimed at performance or memory usage or power consumption or whatever). Compilers mix and match as best they can.
That doesn't prove any such thing - and in fact, the statement is wrong. The programmer-visible ISA does have a significant impact on the performance-critical aspects of the architecture. Not boring stuff, like SSE instructions, but important stuff like the guarantees about memory write ordering and things like that. An x86 chip must pretend that each instruction is fully completed before the next one executes, and has to jump through all kinds of complicated hoops to run them out-of-order internally. A POWER4 chip (now old, but it's one I know off the top of my head) typically executes instructions in groups of four or five, with little or no guarantees about the order in which the instructions within a group are executed - no hoop-jumping needed, the ISA presented lets the compiler handle all that. That makes a significant difference - and there's a whole bunch of similar historical rules that weigh down x86 but aren't present in PowerPC. You cannot decode away restrictions like this. Slapping a big pipeline on the front does not help - Intel thought it would, so they built the P4 around that principle, and then AMD gave them a resounding beating because Intel were wrong (possibly their worst design blunder ever). They have abandoned the overly-extended pipeline in the Core architecture for precisely this reason.
The real reason why this doesn't matter is far simpler. Intel and AMD put millions of x86 chips out each year. That lets them scale their production larger. Higher scale of production gives you lower costs - so even with the estimated 20%-30% (popular ballpark figure, nobody's entirely sure) performance loss from the crappy ISA, Intel and AMD get better performance per dollar than the architecturally superior competition. It's pure economics, nothing to do with the actual chips at all. Raw performance is not important; price/performance is what counts.
If anybody ever comes up with an ISA which gives you a fundamental advantage of a factor of 2 or more, this will change dramatically. So far, nobody has managed anything near that - but the research models are getting better all the time. Someday it might happen.
You have to realise that this is not a "computer" in the sense that you think of one.
This is an abnormally large cellphone. Its feature set is roughly equivalent to what you'd find on a modern 'smart' phone.
They don't market it like that because you'd have a hard time justifying giving free cellphones away to kids as a way of improving their lives. They sell it as a 'laptop', even though it really isn't what people think of when they use the term. It could never survive in a free market - this is a monopoly-based device. Like any device targeting a monopoly market, it is less than ideal.
Frankly, I think the OLPC project's vision is rather small. It could have been so much more.
But hey, you know what they say - give a man a fish, and he'll be fed today. Teach a man how to fish, and he'll sit in a boat and drink beer all day. For some reason, people think that is an appropriate objective, and aim no higher.
Buuu! Energy is not measured in volts. Attempts at debunking things while failing to make basic electrical sense are not impressive. Try harder next time.