I don't know why, it works for Microsoft and used to work wonders for IBM and AT&T. In fact, most stores adopt the last two, these days. (I can't remember the last time I saw a friendly salesperson, or merchandise that actually worked to any meaningful standard. Not just physical merchandise, either. It can take 3-5 days for a text message to go from one coast to the other in the US over T-Mobile.) Mind you, that could be why everything has gone to hell in a handbasket.
GR works great at the macroscopic level, but is lousy at the subatomic level, which is pretty much what you'd expect whether space/time was quantized or continuous. The scale of GR is so fantastically large compared to the granularity, that everything is essentially continuous. In GR, space/time is indeed curved and one of the huge problems with QM gravity is that it cannot be reconciled with GR gravity at the present time. The reconciliation of the two is considered essential for a Grand Unified Theory to exist, and a fair chunk of modern physics only works if you assume a GUT does exist. If the GUT's existance is falsified, that's going to cause a lot of headaches.
So far, that's "textbook" stuff. The next question is whether GR precludes quantized space/time in some other way. The answer is probably no. The equation for relativistic mass is M'=M/(1-sqrt(V^2/C^2)). Both rest mass and relativistic mass are quantized, since mass is supplied in quantized units in the form of the Higg's Boson. The only way you can force M' to hold discrete values is if V also can only hold discrete values. V=S/T (S being the symbol used for displacement, which leads me to believe physicists haven't yet grasped the Latin alphabet quite as well as the Greek one), which in turn means S and T must be constrained, or V would be infinitely variable. If V were infinitely variable, you could convert a mass to energy (E=MC^2) where that energy could NOT be converted back into a mass, because you'd have some non-zero amount of energy left over that did not correspond to a valid Higg's Boson.
(The above paragraph makes the assumption that the Higg's Boson does indeed exist. Iff it does not, GR could permit a totally continuous state.)
Now, can GR's gravity be discrete, using just the notion of curvature in space/time? Perhaps. If I had the answer to that one, I'd be 9/10ths of the way to solving the GUT problem and becoming a celebrity. The odds of this happening are (infinity-1):1 against.
Not quite. What they're saying is that space is quantized, which I think was pretty much an inevitable conclusion. Quantum Mechanics has always said that gravity is particle-based (gravitons).
The logic behind space - and therefore time - being quantized is not trivial and it goes well beyond my ability to describe physics, but a hopelessly trivialized attempt to explain follows. It boils down to the fact that the 2nd Law of Thermodynamics prohibits any point in space/time from being in a zero state, which means all points have to be occupied by something. That something cannot have any arbitrary value, QM insists all values are quantized, they're discrete values. A particle on that scale exists simultaneously in every place the probability curve permits, which is why particles are waves and waves are particles. Thus, position is not independent of the particle, but a property contained by it, making it subject to the same rules as all other properties, which means it's quantized. Since there is no point in space/time in which there is not a particle, and since all particles have a quantized position, space/time must also be quantized, or you violate the 2nd Law of Thermodynamics.
Well, yes, it is about time. Especially as the actual protocols needed were defined a long time ago. (To give you a frame of reference, the DoD were releasing Open Source IPSEC implementations in 1997. Ok, that specific protocol wasn't finalized at that point, but that tells you when the Government was sufficiently capable of and expert at encrypting router communications that they'd admit to it.)
That BGP, DNS and other mission-critical protocols aren't secure even twelve years later says a lot for the extreme lethargy at the level of critical infrastructure. Sure, they can't afford to dive straight in, but since when does the DoD release as Open Source their cutting-edge technology? If they were willing to let potential opponents (such as US citizens) have access, you can be certain they were already considering it old-hat.
It follows that they had the means and capability to install highly reliable, strongly encrypted, strongly authenticated router-to-router and DNS-to-DNS communications within the Internet. Of course, by that time the NSF had sold all the US links to Sprint and assorted other scrap-metal merchants, which is presumably why they never bothered.
It also tells me that the corporate sector is incapable of handling such infrastructure, that the "invisible hand" is too busy playing with itself to worry about such things as security and reliability, that those who believed businesses would be safer hands than universities have been shown to be utterly and completely incorrect.
This is not to say the public sector better. The UK's JANET is hardly a paragon of virtue. It turns out that they're all incompetent, but for different reasons. Businesses know better but want your money at no effort on their part, Governments know better but want your souls at no effort on their part.
I'll put this in programming terms, as it's easiest. "Similar" has two major functional programming prototypes - one with two parameters that always returns a boolean, and one with three parameters that always outputs a float. There are OOP variants of these, where one of the parameters is given as the instance on which the operation is being applied.
similar2 would be A is similar to B. similar3 would be A is more similar to B than to C. (This is legit, as you can do a greater than operation on a real.) The OOP version of similar2 might be: Given A, object B is similar.
It makes no sense, however, in a strongly-typed language, to ask if true is greater than false. (Indeed, even in weakly-typed languages, you should never do this comparison, as it's implementation-dependent. One-sided inequalities are also less efficient, even when you can depend on them.) This is generally seen as a coding bug, though you might pass it off as a waste of CPU cycles. It can "work", say in C, but it can't be trusted to work in the same way even on the same system, and takes more compute power even if you could trust it (or modified the compiler to make sure that it could be trusted).
So, now that I've argued that good coding practices apply as much in English as they do in software, can I pull off an encore and show the comparison is on-topic? Well, yes! In this case, I can, because it's bad programming practices that make Windows XP slower on downloads than Ubuntu! The drain on CPU cycles due to poor coding standards is the primary cause of latency on any system. In this particular case, the latency affects Windows XP, but it also applies to the Linux I/O bug that was discussed earlier. If we genuinely wish to see unnecessary latency eliminated from our lives, we must seek to improve our coding standards rather than munge workarounds for the bugs. Linux' I/O isn't perfect, but the truly impressive latency figures show the superb work done on developing high-quality code rather than having layers-upon-layers of hacks and bit-rot.
It won't win any "I Am Linux" awards, but I would argue that communication would be cleaner, with fewer errors, greater bandwidth and lower latency, if people DID think of themselves in terms of Linux - or, at least, more so than in terms of Windows. (Three parameters, so a greater than is valid.)
They actually released some time back (2 years or so) over 500 patents for Open Source projects. Not sure how many have actually been used by Open Source folks, as most of the patents seemed to be hardware-related and, frankly, Open Cores isn't nearly as popular as Sourceforge. But a good effort none-the-less, in that they did at least make an effort. The 3,000, as others have noted, aren't patents but rather unpatentable ideas they can't market but can get cheap publicity with.
"More similar" is such a fascinating phrase. Regardless (or should that be irregardles?), Linux has better algorithms for TCP control - and a much, much wider range. If Web100 is installed (not, in this case, as it's a default install), it is also auto-tuning. Disk access is also important, and again Linux has a superior range of choices for block I/O handling. Whether comparable choices are superior on one OS or another is a different question, because again this is a default install and so all we care about is whether the default choice is a superior algorithm. Then, there is the matter of context switch overhead. Network I/O and disk I/O both use the kernel, but the application would be in user-space, so you've lots of context switches to ferry data between interfaces. (Which is a stupid design, and there is hardware out that avoids it, but if you're using an x86 anyway, why use efficient hardware?) My understanding is that Windows context switches are much more expensive than those in Linux.
So, if you add up all of these factors, it is entirely possible that the choice of OS alone could indeed make a difference as large as is being reported, and that with a bit of tweaking, it should be possible to achieve vastly superior performance yet.
US Common Law (which, incidently, US courts routinely flout) is mereley a watered-down version of the Magna Carta. I am sure you are familiar with the country that originated this document and their stance on the ICJ. Ergo, Common Law concerns, whilst they may be legitimate, are clearly not the reason. I'm not saying that the US had no legitimate concerns. What I am saying is that those concerns were only found and exploited because of extremely illigitimate concerns regarding a need to protect US officials and agencies involved in activities that go well beyond what is considered reasonable conduct by a nation.
All nations, for example, practice black ops - activities they can't possibly admit to. But many such operations, even though they'd be horrible international scandals if discovered, are likely to fall into the category of "reasonable", given the circumstances. The consequences of them being known, however, so vastly outweigh the reasonableness that some secrecy is to be expected. On the flip-side, you had POWs routinely tortured then thrown out of helicopters in Vietnam. That's not reasonable and no sane person could ever consider it so. Secret CIA prisons in Romania and Poland, politically-motivated assassinations of non-combatants and other "interesting" facets of the "war on terror" are also very unlikely to be remotely reasonable. The gunning down of unarmed civilians in Haditha produced about the same response as some of the better-known massacres of unarmed civilians in Vietnam - a lot of tut-tutting with no culpability and no responsibility.
Now, look for a moment at the politically malicious prosecution claim. The UK is the "bad boy" of Europe (and has been for a long time) but has not faced a single malicious prosecution. France assassinated a member of Greenpeace in New Zealand and was caught red-handed doing so, but I'm not seeing anyone use the ICJ to harass them. Austria is largely run by neo-Nazis and ultra right-wing extremists (and even elected a former Nazi official, Kurt Waldeheim, as President), a move that has thoroughly upset many who are left-of-centre, but nobody is using the ICJ to punish them for their right to do what they like within their own borders.
Conclusion? The US is not, and never has, feared malicious prosecution. Indeed, it wouldn't matter even if such a prosecution happened, as extradition trials ensure that those probably innocent (such as the pilot the US harassed over in Britain and falsely accused - likely knowingly - of training the 9/11 pilots) never get extradited when the trial is likely without merit. So a request is sent. The request can be denied, the same way the UK High Court denied the US request for said pilot. It is for the prosecutor to prove reasonable grounds. Extradition trials probably demand higher standards of proof than the trials the US are entirely happy holding to determine if someone should be charged with a serious crime.
Aha! Now we might be onto something. Clearly, safeguards are in place against malicious use of the ICJ, but this proving of reasonable grounds might be extremely dodgy. The photographs from Abu Gharib would likely be considered reasonable grounds, for example, and had those not escaped into the media anyway, you can bet every penny in the bank that the US Government would have done all in its power, come hell or high water, to prevent any knowledge of such abuses leaking out. Were the ICJ prosecutor to go to a US court and demand extradition, offering such photographs as "reasonable grounds", it would be a serious blow to the cover-ups we have seen time and again in Iraq and Afghanistan. Having the mockeries of a court matial, where information releases can be tightly controlled, at least keeps some of the Government-sponsored terrorism out of sight. Were ICJ prosecutors to go publishing those facts in open court, as part of an extradition process, it would be severely damaging to those who have worked hard (like Adm. Poindexter) to prevent responsibility from being placed.
No, it says the Government may not curtail those rights, but says nothing about individuals having those rights. This is why, for example, slave ownership was acceptable for so long. The Government wasn't the body curtailing the civil liberties. It's also why workplaces are quite entitled to censor employees - the workplace is not the Government and so not subject to those restrictions. The comma in the 2nd Amendment is there, for example, specifically because the "right of the individual" is not something the 2nd can cover and therefore it needs to distinguish between the Constitutional restriction on Government and the effect it is intended to produce. The right is the intended end result, sure, but that is merely the expressed desire and not the expressed control.
Yes, and? The ICJ isn't a court that rules on American law and operates on matters outside United States jurisdiction, so isn't a part of the judicial power of the United States. If it isn't a part of the judicial power of the United States, then the clause doesn't apply. This is why the US can have extradition treaties with other countries, even when those countries do not recognize the rulings of the US Supreme Court as superior to their own law, which would seemingly violate that clause. It doesn't violate it, of course, because again such treaties aren't concerned with the judicial power of the United States. The only thing that is covered by that clause is the ability to arrest people, wich is most certainly authorized by the Superior Court and such inferior courts as otherwise have been established and ordained.
Specifically using clauses like that to prevent due process when it is convenient to do so, but equally specifically defining that such clauses don't apply when that is also convenient, THAT is illegal under that very clause, as it is usurping the role of the Supreme Court to hold the judicial power. The President and Congress, by blocking ICJ membership, claimed a judicial right superior to that of the Supreme Court, which your quote expressly forbids.
There are no rights of the individual in the Constitution. It is a legal document drawn up between the States to define what the Federal Government could and could not do. The only rights that exist in it are rights of the Government. Where rights are mentioned at all, such as in the 2nd, they are used only as an illustration of how to establish when the Government has exceeded the authority granted. The Constitution is not a legal document and defines no law.
In the news today, the Pentagon was ordered to free a man they'd accused of being on one of these suppsed foreign battlefields after failing to provide evidence he was even in the country they supposedly arrested him in. Many of those arrested and imprisoned were captured by bounty-hunters who provided such damning evidence as the culprit wearing a Casio watch. Frankly, if you can't provide any evidence that these people WERE on any battlefield anywhere on the planet, legally or otherwise, then you have little justification for calling them "prisoners of war". And if they ARE prisoners of war, then you are legally required to provide them with all the international safeguards and legal rights that are guaranteed to such prisoners. Providing them with neither the rights of POWs -or- the rights of any other recognized classification of prisoner is simply not acceptable.
The Constitution is the only text guaranteed to apply, as it is the legal document governing the Government. The Constitution is not a document describing the rights of the individual, it is a document detailing the limits of Government. Those limits are carefully specified. Sure, there are recognized parameters, such as commercial speech not being protected under the First Ammendment. Some of these parameters may not be strictly legal, but so far very few have ever been overturned. The right of the President to curtail any clause in the Constitution at whim, simply by claiming war powers, is highly dubious and has usually failed in the Supreme Court when tested. Thus, we can reasonably argue that the suspension of Haebus Corpus is illegal, that the Fifth's guarantees against self-incrimination preclude the use of torture, and that the First prohibits the arrest of an individual on suspicion of being unChristian.
Do any of these affect the ability of an army to protect itself? No. Haebus Corpus means you can challenge your arrest. It doesn't prohibit your arrest in the first place. I would consider being on an active battlefield as an active participant very reasonable grounds for search and seizure. Miranda Rights are unnecessary in the case of a POW, as International Law provides that a person is not required to give anything more than sufficient information for the Red Cross to be able to verify their identity (in pop culture, this is usually given as "name, rank and serial number"). In other words, they already know their rights. Now, if you violate those rights, that's another story. That's a no-no. It's also a war crime. That's one reason the US has been hell-bent on not having those people declared POW. It would involve a sizable chunk of US intelligence and the US military having to explain themselves before the courts. This is also why America didn't want to sign up for the ICJ. It has nothing to do with protecting Americans from false claims, and everything to do with protecting high-ranking officials who happen to also be criminals.
Finally, there are those the courts -and the Pentagon- have ruled already have never been involved in subversive activity, terrorism, or other criminal activity. These people are still not permitted into the US, even though the reasons given for their prohibition are rejected by the very same people who are prohibiting them. That makes sense only as a very sick and twisted act of realpolitik, an effort to avoid looking bad to either side of the fence even when those opposing sides are mutually exclusive.
The cost of RAM is not that great, compared to the cost of a high-end motherboard on a good server, and is absolutely insignificant compared to even a single hour of downtime in any kind of datacentre. If you want genuine 5N's reliability or better (and you can go a lot better than that), you want as little strain on mechanical components as you can get. There's little point in, say, using Carrier-Grade Linux if the practical lifetime of the hard drive due to usage means your hardware cannot maintain a comparable level of reliability.
RAM prices matter for home usage, sure, but since when do home users actually have true data servers? (For that matter, when was the last time you used a Carrier-Grade Linux distro at home?) Most home users have one or two computers, but they don't usually designate a box as a NAS. And even then, most home computers these days have at least a gig of RAM. If you generate more than a gig of long-term data per hard disk read on your home machine, you're using it weird.
I remember when that would buy you 60 megabytes! (Hell, I remember when ONE meg drives cost eight times that.)
If you're running a solid server, you know that mechanical devices are (a) slow, and (b) most under strain when doing anything useful, so you tend to avoid using them when at all possible. Servers should do as much as possible via a RAM-based cache -or- use a RAM disk for data that copies to the hard drive only when necessary.
(So long as RAM is battery-backed, even if the machine crashes or the power goes out, you can recover. Your RAM disk needs to be non-volatile just long enough to reboot.)
Better yet, the drive itself would provide such a RAM disk, transparently, so that all you ever saw was a drive that was slow initially but nearly the same speed as RAM for typical usage.
The ideal would be for drives to become truly intelligent devices (ie: have a decent on-board CPU, not HAL-9000 intelligence, which would be bad for data anyway). It would then be possible to have the VFS and whatever underlying FS' you wanted on the drive itself.
Disk-to-disk operations would then bypass the kernel and asynchronous I/O would consume no primary resources. This was fashionable on some systems (most notably drives that used the IEEE 488 bus) in the 70s and was done to some degree with SCSI, but there's really no excuse for not providing such a capability on any modern drive.
I can see that. You'd need very good thresholds to avoid trivially establishing a self-inflicted DDoS. According to the Byzantine General's Problem, in a system with N nodes, (N+1)/2) of those nodes -must- be trustable in order to detect a node that is not. Thus, it would be necessary to establish in advance that the sum total of all compromised nodes PLUS all nodes run with malicious intent PLUS all misconfigured nodes fell below the ((N-1)/2)-M threshold, where M is the number of nodes you expect to be freshly compromised or otherwise "turn traitor" within the response time of the network, in order to be even mathematically possible. That's hard.
This is the absolute upper limit for a threshold in order for the system to function at all. If the threshold is too low, then the noise in the system (natural or artificial) will destabilize the system. You need the threshold for taking any action at all to be high enough that if it is artificially-induced, it is readily characterizable and readily screened-for. This means that low-volume attacks would simply walk past such a system. Which is fine. The outermost defenses should not be concerned with the same sorts of problems that standard system security is intended to deal with. Using the castle wall analogy, the wall isn't intended to stop you from being attacked, it's intended to stop you from being swamped when you are attacked.
The total effort involved in classifying and filtering must also be cost-effective, so the expense of adding dynamic packet classifiers must be less than the cost (in CPU cycles) of protecting systems in other ways -and- the bandwidth saved by such filtering must exceed the bandwidth consumed by the messaging requirements of such a network.
Next, both the response and the change in response in proportion to the change in stimulus must both be negative feedback loops. There must be increased resistance to adding further security, as security increases, and the increase in security must be less with each step.
Finally, the P2P network must have no possibility of flaps, cascading errors, positive feedback loops, over-reporting or other such characteristics. So, no superlinear, polynomial or exponential response curves. It shouldn't even be linear. Double the stimulus should produce much less than double the response, which does violate the biological model I'm basing this on to some degree, but oaks have greater bandwidth to work with. In other words, it must require a greater DDoS attack on a segment of the system to engender a DDoS effect within the system than would be required to achieve an equal DDoS effect on the nodes within the system individually.
You would certainly have to satisfy all of these requirements for a P2P IDS to work effectively. On a corporate network that has multiple points of access to external networks on multiple sites (typically an extranet-type arrangement), you could probably set up something like this without too much difficulty. However, the problem is so small and constrained, you don't need a generic solution, so there are probably better methods anyway.
Now, doing something like this over any significant federated network of networks like the Internet, that would be a much harder problem. It might very well be that there is no solution that meets all of the constraints in such an environment. That would be interesting to determine. Even if the constraints can be satisfied, the enormous problems and potential weaknesses within such a system would make it very hard to implement in a way that was impervious to self-inflicted wounds.
It's not that it's a surprise, it's that it's well within what many corporate execs and financial experts consume regularly. Which, come to think of it, might explain a great deal, like how one IT company in India hallucinated 95% of its finances. (That must've been some really strong coffee...)
Maybe use a similar sort of system as a massively distributed active intrusion detection system. In "real life" (yeah, the outdoor thing), oak trees have chemicals in their leaves which sublime under normal atmospheric conditions. If the leaves are attacked by insects, the chemicals are released into the air and are picked up by nearby oaks. These respond by adding extra tannin in their own leaves. The practical upshot is that the heavier the insect attack, the more heavily protected the trees become, making it impossible for many insects to attack groves of oaks except en-masse or with powerful acids.
In the digital world, you might use a P2P network as follows: If you get below some threshold of attacks of type XYZ being reported, firewalls might log packets that have similar characteristics as XYZ. If those attacks exceed some minimum threshold and have been reported from some minimum number of sites, you might take more action, such as passing round fingerprints of the attacks and block on those fingerprints being detected. If the attacks pass some higher threshold, you then close suspect ports or re-direct them to honeypots.
Just out of curiosity... if you can image specific viruses in a sample of, say, blood, then would it be possible to do extremely reliable blood screening of any and all known viruses by matching the reconstituted image of each object (or a suitably long cryptographic hash thereof) against a database of known viruses? One of the problems with identifying specific viral strains seems to be that it takes an extremely long time, often relies on the detection of the antibodies rather than the viruses themselves (which doesn't seem reliable, and has produced many stories purporting specific people are immune to viruses those people subsequently die from), and just seems to be all-round a really bad idea if you can avoid it.
If this MRI scanner can image a virus in 3D in such a way that modeling software (possibly with human aid) can isolate the virus and perform a direct match-up, it would seem that you should be able to screen for the virus (a) long before the immune system has detected anything, and therefore before an immune response has occurred, and (b) much more reliably (it should be hard to fool an MRI at this resolution, whereas bacteria and viruses fool the immune system all the time).
However, this line of reasoning is highly dependent on the assumption that the imaging can produce some images that are detailed enough that you can classify things automatically or semi-automatically, but not so detailed that "noise" (irrelevant variations) prevent identification. It is also highly dependent on the assumption that by "3D imaging", they mean images where the component objects can be separated in 3D space, rather than being a 3D "soup" that a trained expert can (eventually) pick objects out from. There are probably other assumptions in there that seem so obvious I'm missing that they're even there.
Even if this is possible (which is a big if), there are any number of factors (price, portability, energy costs, robustness, ease of maintenance, etc) which might preclude it from being used in practice in this way, except perhaps in the case of something particularly dangerous and new. I imagine SARS might have taken less time to classify, had this been available at the time of the initial breakout, assuming I'm correct about how good it is. HIV took even longer to identify, and the cause of CJDnv is only thought to be prions, there is no direct evidence of this - evidence this scanner could possibly provide. (Along, maybe, with an explanation for the other new CJD strain that is affecting Americans.)
Slashdot Mirror
Admiral: Well, at least it's not SMTP.
I don't know why, it works for Microsoft and used to work wonders for IBM and AT&T. In fact, most stores adopt the last two, these days. (I can't remember the last time I saw a friendly salesperson, or merchandise that actually worked to any meaningful standard. Not just physical merchandise, either. It can take 3-5 days for a text message to go from one coast to the other in the US over T-Mobile.) Mind you, that could be why everything has gone to hell in a handbasket.
GR works great at the macroscopic level, but is lousy at the subatomic level, which is pretty much what you'd expect whether space/time was quantized or continuous. The scale of GR is so fantastically large compared to the granularity, that everything is essentially continuous. In GR, space/time is indeed curved and one of the huge problems with QM gravity is that it cannot be reconciled with GR gravity at the present time. The reconciliation of the two is considered essential for a Grand Unified Theory to exist, and a fair chunk of modern physics only works if you assume a GUT does exist. If the GUT's existance is falsified, that's going to cause a lot of headaches.
So far, that's "textbook" stuff. The next question is whether GR precludes quantized space/time in some other way. The answer is probably no. The equation for relativistic mass is M'=M/(1-sqrt(V^2/C^2)). Both rest mass and relativistic mass are quantized, since mass is supplied in quantized units in the form of the Higg's Boson. The only way you can force M' to hold discrete values is if V also can only hold discrete values. V=S/T (S being the symbol used for displacement, which leads me to believe physicists haven't yet grasped the Latin alphabet quite as well as the Greek one), which in turn means S and T must be constrained, or V would be infinitely variable. If V were infinitely variable, you could convert a mass to energy (E=MC^2) where that energy could NOT be converted back into a mass, because you'd have some non-zero amount of energy left over that did not correspond to a valid Higg's Boson.
(The above paragraph makes the assumption that the Higg's Boson does indeed exist. Iff it does not, GR could permit a totally continuous state.)
Now, can GR's gravity be discrete, using just the notion of curvature in space/time? Perhaps. If I had the answer to that one, I'd be 9/10ths of the way to solving the GUT problem and becoming a celebrity. The odds of this happening are (infinity-1):1 against.
Not quite. What they're saying is that space is quantized, which I think was pretty much an inevitable conclusion. Quantum Mechanics has always said that gravity is particle-based (gravitons).
The logic behind space - and therefore time - being quantized is not trivial and it goes well beyond my ability to describe physics, but a hopelessly trivialized attempt to explain follows. It boils down to the fact that the 2nd Law of Thermodynamics prohibits any point in space/time from being in a zero state, which means all points have to be occupied by something. That something cannot have any arbitrary value, QM insists all values are quantized, they're discrete values. A particle on that scale exists simultaneously in every place the probability curve permits, which is why particles are waves and waves are particles. Thus, position is not independent of the particle, but a property contained by it, making it subject to the same rules as all other properties, which means it's quantized. Since there is no point in space/time in which there is not a particle, and since all particles have a quantized position, space/time must also be quantized, or you violate the 2nd Law of Thermodynamics.
No, you'll just hallucinate that you can remember the awesome hallucinations.
Well, yes, it is about time. Especially as the actual protocols needed were defined a long time ago. (To give you a frame of reference, the DoD were releasing Open Source IPSEC implementations in 1997. Ok, that specific protocol wasn't finalized at that point, but that tells you when the Government was sufficiently capable of and expert at encrypting router communications that they'd admit to it.)
That BGP, DNS and other mission-critical protocols aren't secure even twelve years later says a lot for the extreme lethargy at the level of critical infrastructure. Sure, they can't afford to dive straight in, but since when does the DoD release as Open Source their cutting-edge technology? If they were willing to let potential opponents (such as US citizens) have access, you can be certain they were already considering it old-hat.
It follows that they had the means and capability to install highly reliable, strongly encrypted, strongly authenticated router-to-router and DNS-to-DNS communications within the Internet. Of course, by that time the NSF had sold all the US links to Sprint and assorted other scrap-metal merchants, which is presumably why they never bothered.
It also tells me that the corporate sector is incapable of handling such infrastructure, that the "invisible hand" is too busy playing with itself to worry about such things as security and reliability, that those who believed businesses would be safer hands than universities have been shown to be utterly and completely incorrect.
This is not to say the public sector better. The UK's JANET is hardly a paragon of virtue. It turns out that they're all incompetent, but for different reasons. Businesses know better but want your money at no effort on their part, Governments know better but want your souls at no effort on their part.
I was beginning to wonder if the IETF had been bought up or kidnapped by the DHS. That would explain where this "plot" business comes in, anyway.
I'll put this in programming terms, as it's easiest. "Similar" has two major functional programming prototypes - one with two parameters that always returns a boolean, and one with three parameters that always outputs a float. There are OOP variants of these, where one of the parameters is given as the instance on which the operation is being applied.
similar2 would be A is similar to B. similar3 would be A is more similar to B than to C. (This is legit, as you can do a greater than operation on a real.) The OOP version of similar2 might be: Given A, object B is similar.
It makes no sense, however, in a strongly-typed language, to ask if true is greater than false. (Indeed, even in weakly-typed languages, you should never do this comparison, as it's implementation-dependent. One-sided inequalities are also less efficient, even when you can depend on them.) This is generally seen as a coding bug, though you might pass it off as a waste of CPU cycles. It can "work", say in C, but it can't be trusted to work in the same way even on the same system, and takes more compute power even if you could trust it (or modified the compiler to make sure that it could be trusted).
So, now that I've argued that good coding practices apply as much in English as they do in software, can I pull off an encore and show the comparison is on-topic? Well, yes! In this case, I can, because it's bad programming practices that make Windows XP slower on downloads than Ubuntu! The drain on CPU cycles due to poor coding standards is the primary cause of latency on any system. In this particular case, the latency affects Windows XP, but it also applies to the Linux I/O bug that was discussed earlier. If we genuinely wish to see unnecessary latency eliminated from our lives, we must seek to improve our coding standards rather than munge workarounds for the bugs. Linux' I/O isn't perfect, but the truly impressive latency figures show the superb work done on developing high-quality code rather than having layers-upon-layers of hacks and bit-rot.
It won't win any "I Am Linux" awards, but I would argue that communication would be cleaner, with fewer errors, greater bandwidth and lower latency, if people DID think of themselves in terms of Linux - or, at least, more so than in terms of Windows. (Three parameters, so a greater than is valid.)
The Pentium I wishes to disagree.
They actually released some time back (2 years or so) over 500 patents for Open Source projects. Not sure how many have actually been used by Open Source folks, as most of the patents seemed to be hardware-related and, frankly, Open Cores isn't nearly as popular as Sourceforge. But a good effort none-the-less, in that they did at least make an effort. The 3,000, as others have noted, aren't patents but rather unpatentable ideas they can't market but can get cheap publicity with.
"More similar" is such a fascinating phrase. Regardless (or should that be irregardles?), Linux has better algorithms for TCP control - and a much, much wider range. If Web100 is installed (not, in this case, as it's a default install), it is also auto-tuning. Disk access is also important, and again Linux has a superior range of choices for block I/O handling. Whether comparable choices are superior on one OS or another is a different question, because again this is a default install and so all we care about is whether the default choice is a superior algorithm. Then, there is the matter of context switch overhead. Network I/O and disk I/O both use the kernel, but the application would be in user-space, so you've lots of context switches to ferry data between interfaces. (Which is a stupid design, and there is hardware out that avoids it, but if you're using an x86 anyway, why use efficient hardware?) My understanding is that Windows context switches are much more expensive than those in Linux.
So, if you add up all of these factors, it is entirely possible that the choice of OS alone could indeed make a difference as large as is being reported, and that with a bit of tweaking, it should be possible to achieve vastly superior performance yet.
US Common Law (which, incidently, US courts routinely flout) is mereley a watered-down version of the Magna Carta. I am sure you are familiar with the country that originated this document and their stance on the ICJ. Ergo, Common Law concerns, whilst they may be legitimate, are clearly not the reason. I'm not saying that the US had no legitimate concerns. What I am saying is that those concerns were only found and exploited because of extremely illigitimate concerns regarding a need to protect US officials and agencies involved in activities that go well beyond what is considered reasonable conduct by a nation.
All nations, for example, practice black ops - activities they can't possibly admit to. But many such operations, even though they'd be horrible international scandals if discovered, are likely to fall into the category of "reasonable", given the circumstances. The consequences of them being known, however, so vastly outweigh the reasonableness that some secrecy is to be expected. On the flip-side, you had POWs routinely tortured then thrown out of helicopters in Vietnam. That's not reasonable and no sane person could ever consider it so. Secret CIA prisons in Romania and Poland, politically-motivated assassinations of non-combatants and other "interesting" facets of the "war on terror" are also very unlikely to be remotely reasonable. The gunning down of unarmed civilians in Haditha produced about the same response as some of the better-known massacres of unarmed civilians in Vietnam - a lot of tut-tutting with no culpability and no responsibility.
Now, look for a moment at the politically malicious prosecution claim. The UK is the "bad boy" of Europe (and has been for a long time) but has not faced a single malicious prosecution. France assassinated a member of Greenpeace in New Zealand and was caught red-handed doing so, but I'm not seeing anyone use the ICJ to harass them. Austria is largely run by neo-Nazis and ultra right-wing extremists (and even elected a former Nazi official, Kurt Waldeheim, as President), a move that has thoroughly upset many who are left-of-centre, but nobody is using the ICJ to punish them for their right to do what they like within their own borders.
Conclusion? The US is not, and never has, feared malicious prosecution. Indeed, it wouldn't matter even if such a prosecution happened, as extradition trials ensure that those probably innocent (such as the pilot the US harassed over in Britain and falsely accused - likely knowingly - of training the 9/11 pilots) never get extradited when the trial is likely without merit. So a request is sent. The request can be denied, the same way the UK High Court denied the US request for said pilot. It is for the prosecutor to prove reasonable grounds. Extradition trials probably demand higher standards of proof than the trials the US are entirely happy holding to determine if someone should be charged with a serious crime.
Aha! Now we might be onto something. Clearly, safeguards are in place against malicious use of the ICJ, but this proving of reasonable grounds might be extremely dodgy. The photographs from Abu Gharib would likely be considered reasonable grounds, for example, and had those not escaped into the media anyway, you can bet every penny in the bank that the US Government would have done all in its power, come hell or high water, to prevent any knowledge of such abuses leaking out. Were the ICJ prosecutor to go to a US court and demand extradition, offering such photographs as "reasonable grounds", it would be a serious blow to the cover-ups we have seen time and again in Iraq and Afghanistan. Having the mockeries of a court matial, where information releases can be tightly controlled, at least keeps some of the Government-sponsored terrorism out of sight. Were ICJ prosecutors to go publishing those facts in open court, as part of an extradition process, it would be severely damaging to those who have worked hard (like Adm. Poindexter) to prevent responsibility from being placed.
It is no coincidence t
No, it says the Government may not curtail those rights, but says nothing about individuals having those rights. This is why, for example, slave ownership was acceptable for so long. The Government wasn't the body curtailing the civil liberties. It's also why workplaces are quite entitled to censor employees - the workplace is not the Government and so not subject to those restrictions. The comma in the 2nd Amendment is there, for example, specifically because the "right of the individual" is not something the 2nd can cover and therefore it needs to distinguish between the Constitutional restriction on Government and the effect it is intended to produce. The right is the intended end result, sure, but that is merely the expressed desire and not the expressed control.
Yes, and? The ICJ isn't a court that rules on American law and operates on matters outside United States jurisdiction, so isn't a part of the judicial power of the United States. If it isn't a part of the judicial power of the United States, then the clause doesn't apply. This is why the US can have extradition treaties with other countries, even when those countries do not recognize the rulings of the US Supreme Court as superior to their own law, which would seemingly violate that clause. It doesn't violate it, of course, because again such treaties aren't concerned with the judicial power of the United States. The only thing that is covered by that clause is the ability to arrest people, wich is most certainly authorized by the Superior Court and such inferior courts as otherwise have been established and ordained.
Specifically using clauses like that to prevent due process when it is convenient to do so, but equally specifically defining that such clauses don't apply when that is also convenient, THAT is illegal under that very clause, as it is usurping the role of the Supreme Court to hold the judicial power. The President and Congress, by blocking ICJ membership, claimed a judicial right superior to that of the Supreme Court, which your quote expressly forbids.
We did. It's called the IETF. They did. It's called IPSec. Now what's your problem again?
There are no rights of the individual in the Constitution. It is a legal document drawn up between the States to define what the Federal Government could and could not do. The only rights that exist in it are rights of the Government. Where rights are mentioned at all, such as in the 2nd, they are used only as an illustration of how to establish when the Government has exceeded the authority granted. The Constitution is not a legal document and defines no law.
In the news today, the Pentagon was ordered to free a man they'd accused of being on one of these suppsed foreign battlefields after failing to provide evidence he was even in the country they supposedly arrested him in. Many of those arrested and imprisoned were captured by bounty-hunters who provided such damning evidence as the culprit wearing a Casio watch. Frankly, if you can't provide any evidence that these people WERE on any battlefield anywhere on the planet, legally or otherwise, then you have little justification for calling them "prisoners of war". And if they ARE prisoners of war, then you are legally required to provide them with all the international safeguards and legal rights that are guaranteed to such prisoners. Providing them with neither the rights of POWs -or- the rights of any other recognized classification of prisoner is simply not acceptable.
The Constitution is the only text guaranteed to apply, as it is the legal document governing the Government. The Constitution is not a document describing the rights of the individual, it is a document detailing the limits of Government. Those limits are carefully specified. Sure, there are recognized parameters, such as commercial speech not being protected under the First Ammendment. Some of these parameters may not be strictly legal, but so far very few have ever been overturned. The right of the President to curtail any clause in the Constitution at whim, simply by claiming war powers, is highly dubious and has usually failed in the Supreme Court when tested. Thus, we can reasonably argue that the suspension of Haebus Corpus is illegal, that the Fifth's guarantees against self-incrimination preclude the use of torture, and that the First prohibits the arrest of an individual on suspicion of being unChristian.
Do any of these affect the ability of an army to protect itself? No. Haebus Corpus means you can challenge your arrest. It doesn't prohibit your arrest in the first place. I would consider being on an active battlefield as an active participant very reasonable grounds for search and seizure. Miranda Rights are unnecessary in the case of a POW, as International Law provides that a person is not required to give anything more than sufficient information for the Red Cross to be able to verify their identity (in pop culture, this is usually given as "name, rank and serial number"). In other words, they already know their rights. Now, if you violate those rights, that's another story. That's a no-no. It's also a war crime. That's one reason the US has been hell-bent on not having those people declared POW. It would involve a sizable chunk of US intelligence and the US military having to explain themselves before the courts. This is also why America didn't want to sign up for the ICJ. It has nothing to do with protecting Americans from false claims, and everything to do with protecting high-ranking officials who happen to also be criminals.
Finally, there are those the courts -and the Pentagon- have ruled already have never been involved in subversive activity, terrorism, or other criminal activity. These people are still not permitted into the US, even though the reasons given for their prohibition are rejected by the very same people who are prohibiting them. That makes sense only as a very sick and twisted act of realpolitik, an effort to avoid looking bad to either side of the fence even when those opposing sides are mutually exclusive.
The even more tragic thing is that most people probably got the ICHC reference. Hoomins!
The cost of RAM is not that great, compared to the cost of a high-end motherboard on a good server, and is absolutely insignificant compared to even a single hour of downtime in any kind of datacentre. If you want genuine 5N's reliability or better (and you can go a lot better than that), you want as little strain on mechanical components as you can get. There's little point in, say, using Carrier-Grade Linux if the practical lifetime of the hard drive due to usage means your hardware cannot maintain a comparable level of reliability.
RAM prices matter for home usage, sure, but since when do home users actually have true data servers? (For that matter, when was the last time you used a Carrier-Grade Linux distro at home?) Most home users have one or two computers, but they don't usually designate a box as a NAS. And even then, most home computers these days have at least a gig of RAM. If you generate more than a gig of long-term data per hard disk read on your home machine, you're using it weird.
Is that why you've been thrown into the closet by the marauding army ants?
I remember when that would buy you 60 megabytes! (Hell, I remember when ONE meg drives cost eight times that.)
If you're running a solid server, you know that mechanical devices are (a) slow, and (b) most under strain when doing anything useful, so you tend to avoid using them when at all possible. Servers should do as much as possible via a RAM-based cache -or- use a RAM disk for data that copies to the hard drive only when necessary.
(So long as RAM is battery-backed, even if the machine crashes or the power goes out, you can recover. Your RAM disk needs to be non-volatile just long enough to reboot.)
Better yet, the drive itself would provide such a RAM disk, transparently, so that all you ever saw was a drive that was slow initially but nearly the same speed as RAM for typical usage.
The ideal would be for drives to become truly intelligent devices (ie: have a decent on-board CPU, not HAL-9000 intelligence, which would be bad for data anyway). It would then be possible to have the VFS and whatever underlying FS' you wanted on the drive itself.
Disk-to-disk operations would then bypass the kernel and asynchronous I/O would consume no primary resources. This was fashionable on some systems (most notably drives that used the IEEE 488 bus) in the 70s and was done to some degree with SCSI, but there's really no excuse for not providing such a capability on any modern drive.
I can see that. You'd need very good thresholds to avoid trivially establishing a self-inflicted DDoS. According to the Byzantine General's Problem, in a system with N nodes, (N+1)/2) of those nodes -must- be trustable in order to detect a node that is not. Thus, it would be necessary to establish in advance that the sum total of all compromised nodes PLUS all nodes run with malicious intent PLUS all misconfigured nodes fell below the ((N-1)/2)-M threshold, where M is the number of nodes you expect to be freshly compromised or otherwise "turn traitor" within the response time of the network, in order to be even mathematically possible. That's hard.
This is the absolute upper limit for a threshold in order for the system to function at all. If the threshold is too low, then the noise in the system (natural or artificial) will destabilize the system. You need the threshold for taking any action at all to be high enough that if it is artificially-induced, it is readily characterizable and readily screened-for. This means that low-volume attacks would simply walk past such a system. Which is fine. The outermost defenses should not be concerned with the same sorts of problems that standard system security is intended to deal with. Using the castle wall analogy, the wall isn't intended to stop you from being attacked, it's intended to stop you from being swamped when you are attacked.
The total effort involved in classifying and filtering must also be cost-effective, so the expense of adding dynamic packet classifiers must be less than the cost (in CPU cycles) of protecting systems in other ways -and- the bandwidth saved by such filtering must exceed the bandwidth consumed by the messaging requirements of such a network.
Next, both the response and the change in response in proportion to the change in stimulus must both be negative feedback loops. There must be increased resistance to adding further security, as security increases, and the increase in security must be less with each step.
Finally, the P2P network must have no possibility of flaps, cascading errors, positive feedback loops, over-reporting or other such characteristics. So, no superlinear, polynomial or exponential response curves. It shouldn't even be linear. Double the stimulus should produce much less than double the response, which does violate the biological model I'm basing this on to some degree, but oaks have greater bandwidth to work with. In other words, it must require a greater DDoS attack on a segment of the system to engender a DDoS effect within the system than would be required to achieve an equal DDoS effect on the nodes within the system individually.
You would certainly have to satisfy all of these requirements for a P2P IDS to work effectively. On a corporate network that has multiple points of access to external networks on multiple sites (typically an extranet-type arrangement), you could probably set up something like this without too much difficulty. However, the problem is so small and constrained, you don't need a generic solution, so there are probably better methods anyway.
Now, doing something like this over any significant federated network of networks like the Internet, that would be a much harder problem. It might very well be that there is no solution that meets all of the constraints in such an environment. That would be interesting to determine. Even if the constraints can be satisfied, the enormous problems and potential weaknesses within such a system would make it very hard to implement in a way that was impervious to self-inflicted wounds.
It's not that it's a surprise, it's that it's well within what many corporate execs and financial experts consume regularly. Which, come to think of it, might explain a great deal, like how one IT company in India hallucinated 95% of its finances. (That must've been some really strong coffee...)
Maybe use a similar sort of system as a massively distributed active intrusion detection system. In "real life" (yeah, the outdoor thing), oak trees have chemicals in their leaves which sublime under normal atmospheric conditions. If the leaves are attacked by insects, the chemicals are released into the air and are picked up by nearby oaks. These respond by adding extra tannin in their own leaves. The practical upshot is that the heavier the insect attack, the more heavily protected the trees become, making it impossible for many insects to attack groves of oaks except en-masse or with powerful acids.
In the digital world, you might use a P2P network as follows: If you get below some threshold of attacks of type XYZ being reported, firewalls might log packets that have similar characteristics as XYZ. If those attacks exceed some minimum threshold and have been reported from some minimum number of sites, you might take more action, such as passing round fingerprints of the attacks and block on those fingerprints being detected. If the attacks pass some higher threshold, you then close suspect ports or re-direct them to honeypots.
Just out of curiosity... if you can image specific viruses in a sample of, say, blood, then would it be possible to do extremely reliable blood screening of any and all known viruses by matching the reconstituted image of each object (or a suitably long cryptographic hash thereof) against a database of known viruses? One of the problems with identifying specific viral strains seems to be that it takes an extremely long time, often relies on the detection of the antibodies rather than the viruses themselves (which doesn't seem reliable, and has produced many stories purporting specific people are immune to viruses those people subsequently die from), and just seems to be all-round a really bad idea if you can avoid it.
If this MRI scanner can image a virus in 3D in such a way that modeling software (possibly with human aid) can isolate the virus and perform a direct match-up, it would seem that you should be able to screen for the virus (a) long before the immune system has detected anything, and therefore before an immune response has occurred, and (b) much more reliably (it should be hard to fool an MRI at this resolution, whereas bacteria and viruses fool the immune system all the time).
However, this line of reasoning is highly dependent on the assumption that the imaging can produce some images that are detailed enough that you can classify things automatically or semi-automatically, but not so detailed that "noise" (irrelevant variations) prevent identification. It is also highly dependent on the assumption that by "3D imaging", they mean images where the component objects can be separated in 3D space, rather than being a 3D "soup" that a trained expert can (eventually) pick objects out from. There are probably other assumptions in there that seem so obvious I'm missing that they're even there.
Even if this is possible (which is a big if), there are any number of factors (price, portability, energy costs, robustness, ease of maintenance, etc) which might preclude it from being used in practice in this way, except perhaps in the case of something particularly dangerous and new. I imagine SARS might have taken less time to classify, had this been available at the time of the initial breakout, assuming I'm correct about how good it is. HIV took even longer to identify, and the cause of CJDnv is only thought to be prions, there is no direct evidence of this - evidence this scanner could possibly provide. (Along, maybe, with an explanation for the other new CJD strain that is affecting Americans.)