Yes, I'm aware that the Bundeswehr != Wehrmacht, and that in fact soldiers receive specific instruction on disobeying illegal orders. but I've seen the mentality persist in german society. There is a strong presumption of
legality and correctness attributed to authorities. The American skepticism and distrust of authority is hardly present at all, certainly not in any majority.
Well, yes it does. Phishing is fraud irrespective of what is done with the information. Unauthorized access to a computer system is a crime on it's own.
This highlights an extremely important lesson I'd hope West Point and Annapolis cadets learn: Orders _aren't_ Orders!
The US isn't the German "Befehl ist Befehl". A US officer must not blindly obey orders, but has a duty to first determine if the orders are authentic (they weren't, and probably proveably so from the headers), _and_ whether they're legal.
In this case, I would expect a colonel to trust his officers enough to tell them "I'm sending this autoinstal to you". Or his officers to reply "Sir, you sent us an autoinstall without mentioning it. Please confirm this was your intent."
For more than just phishing, there is a temptation to play the Black Hat for user education.
The problem is: "Two wrongs don't make a right". The "education" still involves exactly the same crime as a real exploit. Rather like stealing something a friend had poorly guarded, then giving it back.
Worms like this one just make the case for internal firewalls. Instead of "mostly open" internally, routers can be set to only pass certain ports. If a corp still wants to be mostly open, it can set up tripwires and even autoport shutdowns.
The quote is indeed Einstein. "The Universe is not only stranger than we imagine, it is stranger than we can imagine" [Eddington].
Pgms have indeed seen strange input. Worse since the invention of cut'n'paste.
But I don't think AI need handle all input, so long as it fails gracefully when out-of-bounds. I mostly object to ungraceful failure, especially failure disguised-as-success.
Natural stupidity beats artificial intelligence every time
You cannot make something idiot-proof because idiots are too ingenious (variable). Unfortunately, I see much of AI
as trying to impose order on chaos which cannot be done with deterministic methods. AI _can_ help with data reduction, but not understanding.
I believe a brief exposure to common food-preparation microwaves will fry any RFID device. Even if attached to something metal like a licence plate unless full Faraday-caged.
BTW: barcodes are most easily rendered unreadable by line-color marks parallel and over the lines. One is usually sufficient. More may be necessary for checksumed codes. Better grocery stores do this when they put dated goods on sale. Magstripes are easily destroyed by wiping with rare-earth magnets (check the positioner on an old Hard drive).
Also BTW: an empty wallet is not good when confronted by muggers. They tend to get upset over a poor haul, and commit further violence. Our security people recommend a "bait" wallet with some money, expired credit cards, etc.
People have a right to live how they wish, but any and all security precautions cost. The question is: do they payout? who are the opponents, and how likely are they to act? Personally, I believe that paranoia is a form of egotism: an exaggerated sense of self-importance.
After basically ignoring the SCO UNIX market and worse -- attacking some of his best customers, Darl is doing a volte-face. The key question is: WHY?
Is this a concession that attacking Linux (via IBM) has failed? Or a merely a feint? Is he trying to window-dress SCO for some other buyer? Time will tell
Yes, yellow on white _is_ hard to notice. But extra yellow on magenta makes red and on cyan makes green. Much more noticable, especially on solid areas. When we use color lasers, they're often to repro slides with lots of solid.
I think Ayn Rand would have had a lot of trouble figuring out free software and GPL/Linux. Or even why talented people willingly devote so much time to places like USENET and SlashDot.
I'm not saying Objectivisim has no answers, but it does have some work here. As it does with the great altruism of punishing others.
I are one of these much maligned parents. If my 15yo son knows he wants to play Mortal Kombat:Deception with max gore, I'll buy it (his money) and _insist_ he plays. He gets half the purchase price back when he beats the game. The young man is no idiot, nor some precious infant toddler. If he can hack it, he gets it.
I draw the line at his 14yo sister wanting Britanny's Dance Beat or DDR. _Those_ are cruel and unusual:)
I think you should call them and email them the relevant portions of your logs. It might be not even them due to spoofing, but most likely, it is unauthorized use of their machine. They need to know about a wayward employee.
Yes, fighter pilot Gee suits also have constrictors. I believe these are pneumatically operated, relatively high pressure and uncomfortable even for the short periods of activation.
Fighter pilots are exposed to very low ambient pressure (hence need to breath 100% oxygen). I think they use the concept of a skinsuit -- heavy duty tight spandex to reinforce skin and reduce moisture loss.
I see this as particularly suitable for small diameter closures like gloves -- not much force to hold in. Gloves generating 5 psia hoop stress would only need 5-10 lb force per running inch. Not like keeping a blasted helmet on that might take 500 lb! Or waist closure that is even more (they must have gone to wetsuit-style cross-the shoulders zippers.
AFAIK, US spacecraft stopped being 5psia 100% oxygen after the Apollo One fire. The nitrogen in air is needed to quench the adiabatic flame temperature. Otherwise, switches (aluminum & plastic) ignite. Pressure dones't matter (although it increases reaction rate). Concentration does.
An unanswered question is what low levels of HALON would do. These quench the free-radical combustion mechanism at relatively low levels. But are now banned as suspected Ozone depletors.
ISO the Mendeleev stair-step with a block for "Lanthanides/Actinides", I prefer to carry this further, and separate _all_ the transition elements into a block. If you want remove the H&He ears into a small two-elt block. This highlights orbitals and property families without producing an elongated boat (worse if the La/Ac's are expanded in place).
Sondium on silica nicely solves the nasty Hydrogen transport & storage problems. It presents a few of it's own -- quality control and hydration in transport.
A tougher problem is upstream. Making sodium is gross and inefficient. It's done from brine (salt domes) the in old chlor-alkali process with mercury electrodes. This needs more [over]voltage 3.5V IIRC than hydrolysis. It is the voltage, and particularly the overvoltage needed to drive the process at industrial scales, that makes the process inefficient.
Hit the link, and you will be deemed to be willfully attacking a government security computer. Don you tinfoil hats, the black helicopters will descend!:)
Multiple codepaths aren't unusual, particularly in highly [hand] optimized library routines. But they should be selected based on CPU features returned from CPUID (eax=1) in register EDX, not mfr strings. They have to be anyways, because different "GenuineIntel" have different optimum codings dependant on feature set.
Even if true, this complaint probably isn't worth much. AFAIK, Very little commercial software is prepared with the Intel compiler/libs.
I think someone need to understand a bit better about the purpose of punishment in the legal system. Harsher punishments merely make the system more capricious and arbitrary. "Hung for a lamb, hung for a sheep" escalation. Give police and procesutors more discretion. Hardly something I'd expect the NYTimes to favor.
Perhaps the editurd had an unpleasant episode. He doesn't realize how much his own negligence contributed. Most likely he was using MS software, unpatched at that, and probably even without the NIST recommended configuration [registry].
Those who feel themselves partially responsible often scream blame the loudest. It reduces their own guilt.
Dependancy on compiler instruction level multi-threading compiler technology. Later Alphas are 4-issue RISC. It's hard to keep the threads from stalling. Itanium is EPIC, very much the same thing (6/8 issue?).
Itanium gets _exactly_ the respect it is due. People pay as much attention as they want. Your question really should be phrased: "Why doesn't Itamium get the respect I think it's due?"
That question answers itself: You think differently from most people. Highly specialized, hand optimized massively parallel predictable crunching seems to matter to you. It doesn't to most people. You're in a minority. Get used to it.
BTW, i860 and Alpha suffered from basically the same problem.
I'm hardly surprised at quantum instability. It's inherent in the beast. But it doesn't much matter iff QC has big enough advantages. Put Error Correcting Code on all the busses and storage cells. Plus checkbits on the calcs as AFAIK is done today on P6/K7+ CPUs.
The real question is how deep do you need to make the ECC. That depends on error rate, my guess is Hamming 64+8 ECC will do.
Slashdot Mirror
In this case, I would expect a colonel to trust his officers enough to tell them "I'm sending this autoinstal to you". Or his officers to reply "Sir, you sent us an autoinstall without mentioning it. Please confirm this was your intent."
Pgms have indeed seen strange input. Worse since the invention of cut'n'paste.
But I don't think AI need handle all input, so long as it fails gracefully when out-of-bounds. I mostly object to ungraceful failure, especially failure disguised-as-success.
Natural stupidity beats artificial intelligence every time
You cannot make something idiot-proof because idiots are too ingenious (variable). Unfortunately, I see much of AI as trying to impose order on chaos which cannot be done with deterministic methods. AI _can_ help with data reduction, but not understanding.
BTW: barcodes are most easily rendered unreadable by line-color marks parallel and over the lines. One is usually sufficient. More may be necessary for checksumed codes. Better grocery stores do this when they put dated goods on sale. Magstripes are easily destroyed by wiping with rare-earth magnets (check the positioner on an old Hard drive).
Also BTW: an empty wallet is not good when confronted by muggers. They tend to get upset over a poor haul, and commit further violence. Our security people recommend a "bait" wallet with some money, expired credit cards, etc.
People have a right to live how they wish, but any and all security precautions cost. The question is: do they payout? who are the opponents, and how likely are they to act? Personally, I believe that paranoia is a form of egotism: an exaggerated sense of self-importance.
Is this a concession that attacking Linux (via IBM) has failed? Or a merely a feint? Is he trying to window-dress SCO for some other buyer? Time will tell
I'm not saying Objectivisim has no answers, but it does have some work here. As it does with the great altruism of punishing others.
Millimeter sized? Hell, I'd think a printer was dirty or something. Those are awfully big and noticable.?
I draw the line at his 14yo sister wanting Britanny's Dance Beat or DDR. _Those_ are cruel and unusual :)
I think you should call them and email them the relevant portions of your logs. It might be not even them due to spoofing, but most likely, it is unauthorized use of their machine. They need to know about a wayward employee.
I see this as particularly suitable for small diameter closures like gloves -- not much force to hold in. Gloves generating 5 psia hoop stress would only need 5-10 lb force per running inch. Not like keeping a blasted helmet on that might take 500 lb! Or waist closure that is even more (they must have gone to wetsuit-style cross-the shoulders zippers.
An unanswered question is what low levels of HALON would do. These quench the free-radical combustion mechanism at relatively low levels. But are now banned as suspected Ozone depletors.
A tougher problem is upstream. Making sodium is gross and inefficient. It's done from brine (salt domes) the in old chlor-alkali process with mercury electrodes. This needs more [over]voltage 3.5V IIRC than hydrolysis. It is the voltage, and particularly the overvoltage needed to drive the process at industrial scales, that makes the process inefficient.
Even if true, this complaint probably isn't worth much. AFAIK, Very little commercial software is prepared with the Intel compiler/libs.
Perhaps the editurd had an unpleasant episode. He doesn't realize how much his own negligence contributed. Most likely he was using MS software, unpatched at that, and probably even without the NIST recommended configuration [registry].
Those who feel themselves partially responsible often scream blame the loudest. It reduces their own guilt.
That question answers itself: You think differently from most people. Highly specialized, hand optimized massively parallel predictable crunching seems to matter to you. It doesn't to most people. You're in a minority. Get used to it.
BTW, i860 and Alpha suffered from basically the same problem.
The real question is how deep do you need to make the ECC. That depends on error rate, my guess is Hamming 64+8 ECC will do.