You just need to get the bytes to the destination somehow. It doesn't have to be direct. You could just embed them into an image file and get all your targets to look, or embed them into your secret communications as an anti-interception measure.
If I were a Chinese intelligence expert tasked with meeting this challenge, I'd place my killswitch in the offload engine of network interfaces. Just have to get the chip fabs in China to switch their masks for slightly modified ones, with a tiny bit of extra circuitry on the silicon. It'd look for a specific sequence of 16 bytes in the packet (Putting it in the offload engine ensures it won't inadvertantly break routers en route - at worst you'd knock out a web proxy instead) and, upon detecting them, short every data line on the PCIe interface to ground (or +5v) in the hope of frying the northbridge, or at least crashing the system. Now you've got a simple but effective killswitch. Good for exactly one major use before it's discovered and the trigger blocked, but one use should be quite enough - when the war goes serious, the ability to crash half the US internet will provide many hours of disruption. Enough to cover a first strike. Alternatively, it could be used to quietly fry the webservers of dissidents or proxies - so long as you don't try to hit too many at once, it'd look like nothing more than a failed mainboard and never be detected as a deliberate attack.
You could use it as an ECM system - respond to hacking attempts with a packet containing the kill-code - but if you do that consistantly they'll eventually realise something is going on and start replaying packet dumps until they find the cause.
Christians *today* don't believe those things - but go back a few centuries and you can find plenty of people being executed for violating accepted religious rules and teachings, and a whole series of crusades launched by Christian rulers against Islam (They still dispute who attacked first). Christianity isn't inherently peaceful and tolerant - that is just the form in fashion right now. Maybe in a few more centuries Islam will undergo a peaceful revolution and Christianity will be taken over by theocratic factions, and they can change places.
Either coincidence, or else they reused some old code from another department (As good programmers should - no point spending precious time reinventing the wheel) and thus inadvertantly used the old bug too.
Your feel-good nonsense makes no sense. Why should a 'true' religion teach tolerance? It's actually more internally consistant to preach intolerance: If you accept any religion as true, then by implication all others must be false, and why would a believer tolerate the lies of another faith? Espicially if, as both Christianity and Islam claim, nonbelievers are damned to hell. If you accept that, then tolerance itsself becomes an evil act - condemning others to eternal torment because you are too cowardly to fight the liars on earth.
That just shifts the problem into the design of good electric cars. A problem yet to be properly solved, though they are slowly improving and might be there in another decade or so. Right now they still have severe range problems and take an impractical time to recharge.
It also doesn't help much with energy for transport. Nuclear isn't very portable. You can run a ship on nuclear - many military vessels do just that, but you can't run your car on it.
HTTP caching proxies are somewhat awkward. With a few tweaks, they could be improved.
Firstly, files are identified by address - and even if the file at that address is static, the proxy still needs to do an IMS request with associated 2x-round-trip latency to make sure it hasn't changed. It also means that if the same file is hosted on many different sites, it'll have to be fetched and cached from each one individually. Think of common images and script libraries. There are also security issues - the proxy can easily modify what it sends, which precludes the use of an untrusted proxy server (eg, some prankster might configure it to replace all video files with porn). Not a problem on the corporate lan, a big problem on public hotspots or ad-hoc communications, or when the government might want to sneakily replace images of protesters with happy kittens. With hash-based addressing, files would validate themselves. Just use the already-established SSL to make sure that the HTML pages containing the hashes aren't modified.
HTTP's caching directives help, certainly, but they aren't quite going as far as I'd like.
I know a lot of people who turn off updates, but more because they hate that 'Windows will restart in ten minutes' dialog that always seems to disappear behind other windows and reboot when you're in the middle of a mission on EVE and cost you a hundred-million-ISK battleship.
Fought, and thrashed. The US may struggle to maintain peace and win the support of the locals, but their ability to blow stuff up in conventional war is still quite impressive. It took them under a week to completly destroy the former military and government.
CDNs are nothing new: The big difference is the use of a hash as an address (Which solves security issues of untrusted nodes: Content authenticates itself) and making it public access so all content is treated equally rather than having to pay for hosting.
Government takedown would be possible to some extent, yes. They could easily make a list of 'forbidden hashes' to distribute to ISP-run servers - but with nodes everywhere, they'd never be able to get all copies, and with a hash securing content they couldn't replace with fake. They could render things harder to find (You might have to make tunnels to a few friends) but doing so would be at least as difficult as it is today, if not more so. After all, no central server to shut down.
That only leaves the issue of tracking downloaders.... which would be little more of a problem than it is today, when it's easy to do some logging on an internet connection and see what sites someone accessed - with a little DPI, every site accessed by every user on a major ISP can be logged, and I'm sure China does just that. A problem, but no more of one today. It isn't a revolution that will bring freedom to the internet forever - but it's another tool that'll help.
If nothing else, it'll save a hell of a lot of inter-ISP traffic.
Freenet is slow because it's designed for extreme privacy, not speed. Forget about maintaining anonymisation, and it'd go just as fast as the internet already does at least. You can see that shown already, in torrents - those can easily transfer multi-gigabyte files with great speed. Freenet is just an interesting case study, but think what something like that could do if it didn't have all thse paranoia, and if it was built to follow physical topology.
For example: I want to watch a video on youtube. The laptop doesn't have it cached, so it asks the local network. Fortunatly another family member watched it earlier, and has it - so their computer sends it to mine, rather than using more precious internet capacity to redownload it. Company networks would probably have a node with a few terabytes of storage in for that purpose, to reduce load on the internet connection - it'd work a lot better than HTTP's rather messy if-modified-since method, plus it means most commonly-used resources would still be accessible even if the internet connection or hosting server went down.
Regulation an issue? How about shifting technology in a direction that is harder to regulate? Get ubiquidous encryption going, and someone needs to work on a shift towards a content-addressible network for dissemination. It shouldn't even be difficult.
You could encode CDA addresses as 'HTTP://fallback-http-server/SHA1HASHCAN/hash/mime/mime/filename' - that way you'd have backwards compatibility, as any browsers not programmed to first ask their local CDA cache node if it has that data would fall back to HTTP. Those that are programmed for it would recognise/SHA1HASHCAN as a special pseudo-directory and query their cache, then every open cache on their network before they tried to HTTP it.
CAN is the solution to so many problems. It'd be substantially harder to censor, substantially harder to trace either source or destination of data, eliminate a lot of congestion-causing demand on the internet infrastructure, be more resilient against faults and dramatically reduce the cost of distributing content ensuring that the individuals and small groups on the internet would be just as able to publish large media files as the big boys who can afford global CDNs.
Yes, I'm rather taken with the idea of a distributed, hash-addressible global public cache right now. Storage is dirt cheap, network capacity isn't.
Because an organisation that spends ten years arguing over every diplomatic matter is better than the old-fashioned approach of lobbing shells at each other.
Which raises the question of just how far a search team might go, if they believe the suspect has deliberatly hidden things. A really through search, including things like 'dropped down the U-bend in the bath so it can only be retrieved with a magnet on a stick' wouldn't leave very much of the building intact.
Slashdot Mirror
He never taught them how to use the curses. I'm going to assume there is more to it than knowing the correct words.
You just need to get the bytes to the destination somehow. It doesn't have to be direct. You could just embed them into an image file and get all your targets to look, or embed them into your secret communications as an anti-interception measure.
If I were a Chinese intelligence expert tasked with meeting this challenge, I'd place my killswitch in the offload engine of network interfaces. Just have to get the chip fabs in China to switch their masks for slightly modified ones, with a tiny bit of extra circuitry on the silicon. It'd look for a specific sequence of 16 bytes in the packet (Putting it in the offload engine ensures it won't inadvertantly break routers en route - at worst you'd knock out a web proxy instead) and, upon detecting them, short every data line on the PCIe interface to ground (or +5v) in the hope of frying the northbridge, or at least crashing the system. Now you've got a simple but effective killswitch. Good for exactly one major use before it's discovered and the trigger blocked, but one use should be quite enough - when the war goes serious, the ability to crash half the US internet will provide many hours of disruption. Enough to cover a first strike. Alternatively, it could be used to quietly fry the webservers of dissidents or proxies - so long as you don't try to hit too many at once, it'd look like nothing more than a failed mainboard and never be detected as a deliberate attack.
You could use it as an ECM system - respond to hacking attempts with a packet containing the kill-code - but if you do that consistantly they'll eventually realise something is going on and start replaying packet dumps until they find the cause.
Christians *today* don't believe those things - but go back a few centuries and you can find plenty of people being executed for violating accepted religious rules and teachings, and a whole series of crusades launched by Christian rulers against Islam (They still dispute who attacked first). Christianity isn't inherently peaceful and tolerant - that is just the form in fashion right now. Maybe in a few more centuries Islam will undergo a peaceful revolution and Christianity will be taken over by theocratic factions, and they can change places.
The leap year specification was only written in 1582. So it isn't 2k years old.
Either coincidence, or else they reused some old code from another department (As good programmers should - no point spending precious time reinventing the wheel) and thus inadvertantly used the old bug too.
Your feel-good nonsense makes no sense. Why should a 'true' religion teach tolerance? It's actually more internally consistant to preach intolerance: If you accept any religion as true, then by implication all others must be false, and why would a believer tolerate the lies of another faith? Espicially if, as both Christianity and Islam claim, nonbelievers are damned to hell. If you accept that, then tolerance itsself becomes an evil act - condemning others to eternal torment because you are too cowardly to fight the liars on earth.
That just shifts the problem into the design of good electric cars. A problem yet to be properly solved, though they are slowly improving and might be there in another decade or so. Right now they still have severe range problems and take an impractical time to recharge.
It also doesn't help much with energy for transport. Nuclear isn't very portable. You can run a ship on nuclear - many military vessels do just that, but you can't run your car on it.
At least, I assume it does, otherwise why would the DHS be involved in closing down gambling sites?
Either that, or they are just trying to spend money and justify their existance and vast budget somehow.
Also, first.
That's how NORAD's earthquake/atomic-shockwave-proofing works. The structures are mounted on springs.
The source, of course. In assembler. That's how the first compilers were made, and later rewritten once they were able to compile themselves.
HTTP caching proxies are somewhat awkward. With a few tweaks, they could be improved.
Firstly, files are identified by address - and even if the file at that address is static, the proxy still needs to do an IMS request with associated 2x-round-trip latency to make sure it hasn't changed. It also means that if the same file is hosted on many different sites, it'll have to be fetched and cached from each one individually. Think of common images and script libraries. There are also security issues - the proxy can easily modify what it sends, which precludes the use of an untrusted proxy server (eg, some prankster might configure it to replace all video files with porn). Not a problem on the corporate lan, a big problem on public hotspots or ad-hoc communications, or when the government might want to sneakily replace images of protesters with happy kittens. With hash-based addressing, files would validate themselves. Just use the already-established SSL to make sure that the HTML pages containing the hashes aren't modified.
HTTP's caching directives help, certainly, but they aren't quite going as far as I'd like.
I know a lot of people who turn off updates, but more because they hate that 'Windows will restart in ten minutes' dialog that always seems to disappear behind other windows and reboot when you're in the middle of a mission on EVE and cost you a hundred-million-ISK battleship.
I'm not sure, but I think Apple actually has a patent on that. I remember seeing an image from it a long time ago.
Fought, and thrashed. The US may struggle to maintain peace and win the support of the locals, but their ability to blow stuff up in conventional war is still quite impressive. It took them under a week to completly destroy the former military and government.
CDNs are nothing new: The big difference is the use of a hash as an address (Which solves security issues of untrusted nodes: Content authenticates itself) and making it public access so all content is treated equally rather than having to pay for hosting.
Government takedown would be possible to some extent, yes. They could easily make a list of 'forbidden hashes' to distribute to ISP-run servers - but with nodes everywhere, they'd never be able to get all copies, and with a hash securing content they couldn't replace with fake. They could render things harder to find (You might have to make tunnels to a few friends) but doing so would be at least as difficult as it is today, if not more so. After all, no central server to shut down.
That only leaves the issue of tracking downloaders.... which would be little more of a problem than it is today, when it's easy to do some logging on an internet connection and see what sites someone accessed - with a little DPI, every site accessed by every user on a major ISP can be logged, and I'm sure China does just that. A problem, but no more of one today. It isn't a revolution that will bring freedom to the internet forever - but it's another tool that'll help.
If nothing else, it'll save a hell of a lot of inter-ISP traffic.
I'm sure a lot of the 'free Tibet' campaigners just do it to annoy China.
Freenet is slow because it's designed for extreme privacy, not speed. Forget about maintaining anonymisation, and it'd go just as fast as the internet already does at least. You can see that shown already, in torrents - those can easily transfer multi-gigabyte files with great speed. Freenet is just an interesting case study, but think what something like that could do if it didn't have all thse paranoia, and if it was built to follow physical topology.
For example: I want to watch a video on youtube. The laptop doesn't have it cached, so it asks the local network. Fortunatly another family member watched it earlier, and has it - so their computer sends it to mine, rather than using more precious internet capacity to redownload it. Company networks would probably have a node with a few terabytes of storage in for that purpose, to reduce load on the internet connection - it'd work a lot better than HTTP's rather messy if-modified-since method, plus it means most commonly-used resources would still be accessible even if the internet connection or hosting server went down.
Regulation an issue? How about shifting technology in a direction that is harder to regulate? Get ubiquidous encryption going, and someone needs to work on a shift towards a content-addressible network for dissemination. It shouldn't even be difficult.
/SHA1HASHCAN as a special pseudo-directory and query their cache, then every open cache on their network before they tried to HTTP it.
You could encode CDA addresses as 'HTTP://fallback-http-server/SHA1HASHCAN/hash/mime/mime/filename' - that way you'd have backwards compatibility, as any browsers not programmed to first ask their local CDA cache node if it has that data would fall back to HTTP. Those that are programmed for it would recognise
CAN is the solution to so many problems. It'd be substantially harder to censor, substantially harder to trace either source or destination of data, eliminate a lot of congestion-causing demand on the internet infrastructure, be more resilient against faults and dramatically reduce the cost of distributing content ensuring that the individuals and small groups on the internet would be just as able to publish large media files as the big boys who can afford global CDNs.
Yes, I'm rather taken with the idea of a distributed, hash-addressible global public cache right now. Storage is dirt cheap, network capacity isn't.
Because an organisation that spends ten years arguing over every diplomatic matter is better than the old-fashioned approach of lobbing shells at each other.
One of the worst things about the cloud is that it can go wrong when someone else screws up, so you get the blame for their mistakes.
I find it becomes more blatantly contradictory if you express it as 9/11s-per-month.
Which raises the question of just how far a search team might go, if they believe the suspect has deliberatly hidden things. A really through search, including things like 'dropped down the U-bend in the bath so it can only be retrieved with a magnet on a stick' wouldn't leave very much of the building intact.
Even his correction is wrong, as 'cracker' is also commonly a term for someone who specialises in breaking software copy-prevention measures.