In fact, if anything, Schneier has a conflict of interest in that the less secure the Internet is, the easier it will be for him to sell his services.
OTOH, the more secure the Internet is, the less work Counterpane has to do to provide a particular level of service. It analogous to insurance companies that require certain fire countermeasures as a condition of providing insurance (extinguishers, real firewalls, sprinklers,...). It is not obvious where the line between conflict of interest and public service is drawn though.
I'm not sure about TightVNC, but plain VNC servers under Windows have to repeatedly read out the entire frame buffer from the video card and compare it with the previous frame to detect changes (because they don't/can't hook Windows to find out just which rectangles changed). For some combinations of video card, bus, resolution, and color depth this can eat a lot of CPU time, regardless of how little network bandwidth it uses. A lower color depth or resolution can speed it up.
<unix plug> The X-VNC server fixes this problem by design since it knows exactly which pixels are changed by every graphics operation. </unix plug>
A question: Do you need security? Most VNC servers and clients provide no security. They just broadcast everything you type -- including passwords -- on the network with no encryption. I personally would not use ordinary VNC programs *at all*. If you want to, hey, it's your data.
Cablesonline has the electrical adapter by itself for $8. It'll work if you just need to hook the drive up for a few minutes to suck the data off. Just be carefult not to short out the exposed circuit board on the drive. FWIW, I ordered a buttload of IDE paraphenalia from them back in February and was pleased with the ordering process, merchandise quality, and delivery time. (Which I mention only because some of these online "stores" can't seem to find their asses with both hands, a map, an instructional video, and a 25 hour remedial course in applied ass-finding.) They also have all sorts of ribbon cable, IDC connectors, and gender changers for the 3.5" and 2.5" IDE standards.
Regarding the broken laptop system board, I'd see if somebody could repair it (somebody who actually knows how to use a soldering iron, not one of these swap-out-the-board-depot monkeys (no offense)). Unless the circuit board itself is broken, it's probably salvegable. (And even then, with a little bailing wire and luck.) If it was me, I'd already have the soldering iron warming up, but then I do have a rather hands on approach.
Stocks, bonds, and commodities have long been sold using computers, and real-time confirmation is an essential part of the business. Commodities sales in particular constitute an actual purchase of deliverable physical chattels, and are indistinguishable from modern 'e-commerce'. Ditto for bank transactions that involve currency conversion (which unlike a simple transaction constitute a sale of goods at a profit).
Frankly, I wouldn't lift a damn finger until the patent holder sued for willful criminal infringement, then I'd rape the individuals involved (not the companies -- if they are gonna personally be idiots, they have to be personally punished) violently with prior art and fraud counterclaims. But then I tend to have a scorched earth approach...
Slashdot needs to start selling subscribtions now, and start thinking about ways to sell advertising. Personally, I would subscribe, and I bet most daily readers of this site would.
I would definitely pay a few bucks a month to support Slashdot (as I do for Kuro5hin.org). I find that/. consistently keeps me weeks ahead on the most important tech news topics, and that's worth at least the price of a magazine subscription to me. (Sure, I could read Wired, C-Net, The Register, blah blah fucking blah, but I don't have the time.)
I always thought the VA linux hardware was a good route for them,...
Except that their hardware was rather expensive. I kept checking their prices, but they were always hundreds of dollars too expensive.
IMHO, their mistake was trying to go from a niche market (nice Linux-running boxen) straight to a megacorporation with a wide range of products (a la IBM). They tried to make that jump by maximizing the burn rate, but burn rate can only buy green employees and hardware. It cannot buy an experienced engineering and development staff, mature software products, and all-important customer relationships and business partnerships. It is possible to build a large diverse company, but you have to expand in stages with attention to profitability every step of the way. E.g., like how Microsoft came from no where and dethroned IBM and DEC. The whole 'Instant IBM' approach was just doomed.
If you look at VA, their strategy was 100% Instant IBM. They tried to dominate the hardware market before they had the mature software and hand-holding support to make the extra cost worthwhile. They bought Slashdot to preach to the converted and shill house products. (Remember the Adfu days when/. banners occassionally had interesting products that actual geeks might buy? The only thing even vaguely interesting these days is Think Geek.) They threw *huge* amounts of money at bandwidth, hosting, and server administration in the hope of increasing the amount of free software. Nevermind that VA would have neither licenses nor expertise in the software thus developed, and could therefore not directly profit from it.
So what are they left with? Banner ads (ha!) for things I don't want to buy (ha! ha!) and SourceForge. SourceForge support and cusom development can probably be made profitable by itself -- it's a useful tool -- but even if it is maximally successful and they get a CEO with a winning strategy, it'll take a decade to recover the capital they pissed away. I don't see them getting such a CEO (although the board could surprise us), so don't see VA even being a software powerhouse.
If I was part of the/. crew, I'd be thinking hard about how to turn what they have into something sustainable. Random ideas: banner ads for tech products I might actually buy, paid placement of a few stories a week, paid Slash hosting (product support sites, religious/political sites, government sites), closed-source turn-key Slash installer,...
Also, on linux, I was under the impression that the virtual terminals *were* limited - when you compile a kernel the default is 256 ptys; I may just be utterly wrong here, but I thought that meant there was a 256 virtual terminal limit.
True, although 256 ptys is probably more than enough for any single-user system. If you need more you can build a kernel that supports (IIRC) 2048 ptys.
I would presume it also applies to X - but does X open all of it's stuff in a single console terminal?
Ptys look like a serial port to the downstream process, but upstream is another process instead of serial port hardware. Thus they're only used for things like terminal emulators that really need serial port compatibility. Most X programs don't use them.
I hope so, if only so that the memory and research of Digital Equiptment Corp lives on. (man those guys were cool!)
<sigh> I remember loading http://www.altavista.digital.com/ on an early Netscape. It was *SO* much better than archie. It was under digital.com instead of the familiar altavista.com because it started out as a DEC R&D project and they didn't think to register altavista.com, and some random guy got it instead.
Ever wonder why the States are doing their best to quash independent private efforts to go in Space? Because it knows very well that it will trigger an exodus of the best minds of the planet.
Incidentally, a non-Terrestrial (say, Lunar) nation would be able to bombard Earth with kinetic energy weapons (i.e., rocks the size of buildings moving at 50 kilometers per second), and would also be fairly invulnerable to nuclear bombs (no air means no shock wave and no fallout). Not only would the best people leave, they'd have unilateral assured destruction capability. Yet another reason for groundhog politicos to squash private launch capability. Dammit.
If only the Chinese dissidents at Tiananmen Square had been "packing", those tanks wouldn't have stood a chance.
That statement is absolutely accurate. Undefened tanks (and planes, ships, etc.) are hideously vulnerable. A bunch of primitives can simply walk up to a tank that isn't defended by infantry and jam its treads with sticks and rocks. Then they can build a fire underneath it at their leisure and cook the tank crew to death. All mechanized weapons are similarly vulnerable. Look at what a couple of random attackers were able to do to the U.S.S. Cole with a raft, fertilizer, and oil.
If you can take out the infantry guarding the mechanized weapon, you can destroy the weapon. If you have small arms, you *can* take out the infantry.
I am so tired of this absurd argument by gun nuts that citizens with pistols, rifles, and shotguns can successfully defend themselves against a government gone bad.
The argument, however, is true. If every block of a city is defended by a modest supply of small arms, the city is unconquerable. Destroyable, perhaps, but unconquerable. If there are enough defenders with guns, and the government-gone-bad doesn't have the will to exterminate the city, the revolution is successful.
Tanks, planes, and bombs are relatively immune to some bunch of yahoos with Glocks, Rugers, and Colts.
You can only be conquered if the enemy can send in flesh and blood people to impose their word as law. If you have small arms, you can keep sending those would-be conquerors back home in coffins indefinitely. If the would-be conqueror is not willing to use weapons of mass destruction, they must eventually withdraw.
Small arms also tend to keep the police and other government enforcers reasonable. If John Suspect might be carrying a gun, law enforcement won't be nearly so quick to put him in a position where he has nothing left to lose. Ditto for prospective mass murderers, muggers, rapists, and so forth.
The whole we were placed on these lists was not due to anyone complaining about spam originating or being relayed from our server, but just because it had an open relay.
Shoulda used POP before SMTP: only allow SMTP access to IP addresses that have successfully authenticated with POP in the last few minutes. Since most email programs automatically check for new messages the moment they start, and repeatedly check every few minutes thereafter, legitimate users don't even know you are filtering. To the best of my knowledge, lots of ISPs have good success with this technique.
Don't blame the MAPS/ORBS because you can't deploy a trivial and obvious technical solution that thousands of other people use.
Most of these Blackhole lists do send a message back to the person trying to send the mail, and they often portray admins who run open relays as evil spammers or complete morons. Neither of these is true.
Anyone who operates a high-gain publicly-accessible network data amplifier *IS* either evil or a moron. Smurf amplifier, 0WN3D unix box, open mail relay, who cares. LART 'em till they glow then shoot 'em in the dark.
(Not that I'm not sympathetic to the difficulties of being a small ISP: I just don't think there's any excuse for operating an abusable data amplifier.)
Mail is about the same as Internet from a strict legal point of view.
No. Mail fraud is a specific federal crime.
True, but I was speaking to end results. Blatant fraud--such as failing to deliver on an eBay auction--is pretty much unlawful in every jurisdiction. In my opinion, the useful thing about transactions involving the U.S. mail is not that there are particular laws governing it, but that the US Postal Inspection Service is very proactive about investigating complaints and intimidating offenders.
The thing about it is, mail fraud is a federal crime which much higher penalties than other forms of fraud (Internet fraud being generally unclassified).
Mail is about the same as Internet from a strict legal point of view. The difference is that the Postal Inspectors pay attention to every complaint they receive and they're humorless professional government enforcers with considerable discretionary power. If it goes through the mail, they can kick ass and take names for you. If it doesn't, you are more on your own.
Auction fraud is illegal, I believe there is a way you can file with the FTC and have the fraudelant seller fined thousands of dollars, or even possibly jailed. IIRC, something about interstate commerce, or whatknot.
The US FTC doesn't usually handle individual consumer complaints. However, they are interested in finding patterns of abuse and ripping new assholes for serial criminals. It might not immediately help, but filing a complaint on their web page only takes a moment. The various Better Business Bureaus serve a similar function. You can also send complaints to you state/regional Attorney General.
As I've pointed out, look through a pane of (clean) glass, then through some plexiglass.
That's merely a matter of purity. Low purity anything will suck (e.g., the noticeable green tint of the edge of a typical sheet of window glass). This is high-purity PMMA fiber from Edmund Optics. (PMMA is polymethyl-methacrylate, the fancy chemical name for plexiglas.) A ten meter length of it lets up to 80% of the light through (exact amount depends on color).
I'm not saying that its relevant for audio, sicne I have no clue what more THD a TOSLink cable adds.
That's the beauty of digital: if enough signal gets through, the error rate will be microscopically small. Although the converse is also true: if too little signal gets through, the error rate becomes ludicrously large.
You took it that I recommend glass, or think its stupid to use plastic... on the contrary, I pointed out the irony of using a cheap plastic part in an expensive stereo.
Sorry, I was reading too much into your comment. However, I don't see it as ironic: it's good engineering to use the cheapest thing that gets the job done with plenty of safety margin. If that thing happens to be cheap plastic, so be it. It's really no different than using a $5 Ethernet cable on a $15,000 computer.
Electrons, on the other hand are 'fermions', which means that they interact strongly with other electrons.
Er, not quite. The real issue is that electrons have a charge for a strong, long-range force. Even if you used a superconductor (where the electron pairs are bosons), you'd still have to deal with inductive and capacitive effects, as well as dielectric losses (which are non-negligible at >1GHz for common circuit board materials). Photons are convenient for communications because they are uncharged, not because they are bosons.
Think of the EXTRA heat it would create....thats right EXTRA heat, from the light sources, and for every connection beween 2 points you need 2 light sources. I don't get how this is supposed to help heat problems.
Suppose you have a 256-bit electrical bus, where each wire has 10pF load capacitance, swings between 0V and 3.3V, and toggles at an average rate of 2.5GHz. The aggregate transfer rate will be 80Gbyte/sec. This will be a fairly standard bus situation for a circa 2010 computer.
The power needed to drive that bus will be 1/2 * N * f * C * V^2 = 0.5 * 256 * 2.5*10^9 * 10*10^-12 * 3.3^2 = 35Watts.
35 watts just for interconnect! Even if the optical interfaces consume a whopping 250mW each, you can still afford 140 of them for the power cost of copper.
But wait, there's more: modern CPUs need a huge L2 cache to compensate for the narrow pipe to main RAM. If you widen the pipe, you can get away with a lot less L2 cache, which saves a lot of power (cache is typically static RAM, which has 4 to 6 transistors per bit and sucks a lot of power). Optical interconnect can potentially provide a dedicated link between each RAM unit and the CPU. The latency will probably be higher, which will penalize things like office suites that have a random pattern of memory accesses, but signal processing, graphics, and technical calculations will be blazingly fast.
FWIW, a good portion of the consumer-level TOSLink fibers are made out of plastic, not glass. Kind of funny, mid-level consumers trying to be prosumers by buying all optical/digital stuff then using plastic to connect them.
Total foolishness. The *only* thing that matters for optical communication is whether the signal-to-noise ratio is large enough, which turns into the question "Is this cable's attenuation small enough?" If the answer is "yes", then the cable is totally acceptable. Period. End of story.
If there are several cables that are acceptable, you then ask "Which is cheapest?" and "Which is most durable?" and the answer for those is a resounding "plastic". Frankly, using glass fiber to move an audio signal a couple of meters would be really, really dumb.
This would put SCSI on the skids. Right now SCSI is the only really fast interface commonly available between devices, but it's cost has kept it from becoming the standard. But if you could just plug in a fiber connection, you'd be rocking.
SCSI is rather physical layer agnostic. It already runs on at least four totally different electrical layers: high-volvage single-ended, high-voltage differential, low-voltage differential, fibre channel (which can be copper, despite the name). Optical SCSI would be just another physical layer. The real value of SCSI is that it is very nicely tailored to mass storage devices.
Another problem is that we'd still have the silicon-to-light translation bottleneck. i.e. and electrical signal from a pin on a chip needs to be converted to laserlight somehow. To make this truly work, you'd need a chip that reponds via light,...
Yup, that's the real challenge. Speaking from personal experience with optical chip modules, getting fiber/light to the chip is major pain in the ass. The mechanical design challenges are significant and obnoxious.
But here's an odd question: why don't the Vietnamese hate the US (in general, on average)?
A good question. I don't know a lot about Vietnamese culture, but I suspect there are several factors. For one thing, the US occupied the south without too much razing and pillaging. If US forces had behaved like the Japanese occupation of Nanking the situation would be different. Another big factor is that the government after withdrawal was fairly rational and progressive, unlike Pol Pot or Chairman Mao. A moderate communist government that isn't a blatant tool for Moscow is not such a bad thing. Yet another factor is that they prevailed in battle. There's also the fact that many of the extremists and nutcases got themselves killed.
Oh, we made to say that that was our plan... after the fact.
It was obvious to anyone who visited the Soviet Union that their statements of population, technological level, and industrial capacity were grossly exaggerated, and thus obvious that they wouldn't be able to keep up in a contest of research and production. This was fairly widely recognized, at least among US strategy analysts, from at least the mid '60s, and was widely recognized by the late '70s. The economic/industrial contest was deliberate. So were the interesting diplomatic measures: the State Department would take every visiting Soviet citizen on trips to shopping malls, fast food joints, music stores, and so forth, on the theory that they'd go back home and be rather dissatisfied with the Soviet standard of living.
Apparently you are not very up to date on contemporary guerilla warfare. If you want to see what a small number of psychotic troops fighting for their homeland can do, take a look at the ass-beating the USA took in Vietnam.
Hardly. The Vietnam War was a proxy war between the US and the USSR, an extended campaign in the Cold War. North Vietnam had essentially zero industrial capacity for fighters, bombers, aircraft, firearms, radios, and anti-aircraft missiles. Without vast infusions of Soviet materiel, the US would've promptly conquered North Vietnam. (Of course without the Soviet presence there would have been no need to.)
Even with support from the USSR, the US was winning the war at the time of withdrawal. The withdrawal occurred because of the stunning PR incompetence of the US government. They didn't understand the tremendous power of an extended publicity campaign. They could probably have gotten support for a proxy war against the USSR, but they were silent and all that the public saw was an endless parade of body bags, year after year after year. As there was no strong leadership, the American public grew tired and ended the campaign.
An even better example would be the people of Afghanistan, who... even took on the Soviet Army and beat them back, albeit with a good bit of help from the USA. (Emphasis mine)
Yet another major campaign in the Cold War, again a proxy war in fairly worthless territory, territory that neither nation would have bothered with were it not for the other superpower. Again, the Soviets sent vast amounts of materiel into the theatre, and again the US-supported forces destroyed most of what they sent in. Unlike the Vietnam War, the Soviets also sent lots of soldiers into Afghanistan, which was a lethal US-funded meat grinder.
Something you have to understand about the Soviets was that their technology was not efficient. Compare to US factories, it was much more expensive for them to build a tank or fighter. The effectiveness and quality of Soviet war machines also tended to be rather low. The net result is that it cost the Soviets many more man-hours to field a credible military force. So when they sent in a tank that got promptly bombed by US-supported forces, they had to divert a lot more industrial capacity away from luxury goods, research and development, and so forth. At the same time, they spent far to much of the remaining industrial and R&D capacity trying to outdo the Strategic Defense Initiative. It's also worth pointing out the substantial diversion of Soviet R&D during the Vietnam Proxy War as they tried to compete with the US Apollo project.
Put all this together: previous costly war with no obvious victory and simultaneous loss at a technical competition, currently costly war with no obvious victory and heavy personnel losses and an even bigger unwinnable technological competition. It broke the will of the Soviet government and impoverished the people. The loss in Afghanistan was the straw that broke their back.
40,000 guerrillas in a war torn nation full of refugees are worse than millions of troops fighting with conventional styles of
warfare.
Only with extensive support from a superpower, and that ain't gonna happen for Al Quaida or the Taliban. Especially since the US-aligned Special Operations groups will be using what are, frankly, guerrilla tactics. It's going to be guerrilla versus guerrilla, only the US guerrillas will have C-130s full of materiel arriving as needed, good air support, night-vision scopes, satellite reconnaissance, encrypted spread-spectrum radios, and so forth.
The US also has a major advantage: they are not trying to conquer and hold Afghanistan like the Soviets were. They are simply trying to kill and disrupt a certain few thousand people. Also, unlike Israel or Iran, the only US criterion for an Afghan govt is peacefulness, stability, and cooperation with US intelligence. They won't be trying to prop up a violent government to fight a proxy war against a major power (in fact, they'll be specifically avoiding such a govt).
You think some pistol-packing cowboys would win against a nuclear strike?
Firstly, it wouldn't be cowboys with pistols, it would be hunters with scope-sighted rifles and the patience to wait all day for a target to wander into range. Secondly, the US would never use nuclear weapons in a revolt, because that would just teach the next crop of revolutionaries that mutually-assured destruction was a necessary tactic.
You are saved from US retribution *purely* because it looks bad for the US govt to massacre hundreds of thousands of its own people.
Pure fantasy. History has proven time and time again that even the US govt will kill Americans when provoked.
Terrorism (action against a STATE) is NEVER an acceptable form of political debate. Put down your guns, and talk.
You're *so* right. If only the Jews had *talked* to the Third Reich in WWII....
If you can convince an entire area to get up in arms, surely you could also convince them to march on Capitol hill, or all to work in political activism.
The situation tends to become ripe for revolution gradually, so that the participants don't realize it. In retrospect it might be obvious, but then and there the gradual escalation is imperceptible. By the time it boils over it's too late for politics.
Those times have since given way to civilised living.
Civilization is not a state of being, nor an accomplishment. A society does not go to Civilization High School for a few years and get a Civilization Diploma that proves they are Officially Civilized for all time.
Civilization is a continuous process, and a rather chaotic and poorly controlled one at that. There are human forces continually working for and against it, and occassionally the opposition wins out. The question is not how far can civilization rise, but how far can it fall. Historically, the most spectacular falls -- falls that brought the societies involved to their knees -- have occurred when the populace has become weak and a central imperial power has become strong. Think Imperial Rome shortly before the fall, Imperial Japan in the early 20th century where the overenthusiasm of the powerful military class brought the nation to ruin, Germany during the Third Reich years, or America during the War Between the States. Power disparities invariably breed violent revolution. Conversely, power equality and lack of social classes tends to breed stability, peace, and liberty.
It's the threat to their life (represented by the pointing of a sword/gun at them) that makes them do it.
That was my point. The government always has the sword. The only question is whether they will be able to push you around with it, or whether you can push back.
So where exactly is it in the constitution that you have the right to make threats against another's life?
It's not so much a threat as an insurance policy. Pushing around a citizen might be illegal, but pushing around an armed citizen is unhealthy. A paper justice system is easy to corrrupt, but no number of corrupt judges can sew the ears back on a dirty cop.
When I heard that current technology was easily defeated that matched up with everything I learned in school (graduated in '98). People have been working to make computers identify pictures for a long time.
This is a major fallacy I've been seeing a lot, judging the strategic value of a subject by existing off-the-shelf hardware. Very little work has actually been done on machine vision. If face recognition turns out to be a critical limitation in the war, and the US pulls out all the stops a la the Apollo project and puts twenty or thirty billion dollars into surveillance technology, there will be tremendous advances.
It's probably a poor practice to count on the idea that they are too dumb to wear an ear ring and a beard.
Colors and nearby visual noise are small problems. Facial hair is a bit tougher, but still leaves much of the face uncovered, and terahertz radiation probably goes right through it. (Unconventional illumination is another area that has been almost totally ignored by conventional machine vision.)
Slashdot Mirror
<unix plug> The X-VNC server fixes this problem by design since it knows exactly which pixels are changed by every graphics operation. </unix plug>
A question: Do you need security? Most VNC servers and clients provide no security. They just broadcast everything you type -- including passwords -- on the network with no encryption. I personally would not use ordinary VNC programs *at all*. If you want to, hey, it's your data.
Regarding the broken laptop system board, I'd see if somebody could repair it (somebody who actually knows how to use a soldering iron, not one of these swap-out-the-board-depot monkeys (no offense)). Unless the circuit board itself is broken, it's probably salvegable. (And even then, with a little bailing wire and luck.) If it was me, I'd already have the soldering iron warming up, but then I do have a rather hands on approach.
Frankly, I wouldn't lift a damn finger until the patent holder sued for willful criminal infringement, then I'd rape the individuals involved (not the companies -- if they are gonna personally be idiots, they have to be personally punished) violently with prior art and fraud counterclaims. But then I tend to have a scorched earth approach...
IMHO, their mistake was trying to go from a niche market (nice Linux-running boxen) straight to a megacorporation with a wide range of products (a la IBM). They tried to make that jump by maximizing the burn rate, but burn rate can only buy green employees and hardware. It cannot buy an experienced engineering and development staff, mature software products, and all-important customer relationships and business partnerships. It is possible to build a large diverse company, but you have to expand in stages with attention to profitability every step of the way. E.g., like how Microsoft came from no where and dethroned IBM and DEC. The whole 'Instant IBM' approach was just doomed.
If you look at VA, their strategy was 100% Instant IBM. They tried to dominate the hardware market before they had the mature software and hand-holding support to make the extra cost worthwhile. They bought Slashdot to preach to the converted and shill house products. (Remember the Adfu days when /. banners occassionally had interesting products that actual geeks might buy? The only thing even vaguely interesting these days is Think Geek.) They threw *huge* amounts of money at bandwidth, hosting, and server administration in the hope of increasing the amount of free software. Nevermind that VA would have neither licenses nor expertise in the software thus developed, and could therefore not directly profit from it.
So what are they left with? Banner ads (ha!) for things I don't want to buy (ha! ha!) and SourceForge. SourceForge support and cusom development can probably be made profitable by itself -- it's a useful tool -- but even if it is maximally successful and they get a CEO with a winning strategy, it'll take a decade to recover the capital they pissed away. I don't see them getting such a CEO (although the board could surprise us), so don't see VA even being a software powerhouse.
If I was part of the /. crew, I'd be thinking hard about how to turn what they have into something sustainable. Random ideas: banner ads for tech products I might actually buy, paid placement of a few stories a week, paid Slash hosting (product support sites, religious/political sites, government sites), closed-source turn-key Slash installer, ...
If you can take out the infantry guarding the mechanized weapon, you can destroy the weapon. If you have small arms, you *can* take out the infantry.
The argument, however, is true. If every block of a city is defended by a modest supply of small arms, the city is unconquerable. Destroyable, perhaps, but unconquerable. If there are enough defenders with guns, and the government-gone-bad doesn't have the will to exterminate the city, the revolution is successful. You can only be conquered if the enemy can send in flesh and blood people to impose their word as law. If you have small arms, you can keep sending those would-be conquerors back home in coffins indefinitely. If the would-be conqueror is not willing to use weapons of mass destruction, they must eventually withdraw.Small arms also tend to keep the police and other government enforcers reasonable. If John Suspect might be carrying a gun, law enforcement won't be nearly so quick to put him in a position where he has nothing left to lose. Ditto for prospective mass murderers, muggers, rapists, and so forth.
Don't blame the MAPS/ORBS because you can't deploy a trivial and obvious technical solution that thousands of other people use.
Anyone who operates a high-gain publicly-accessible network data amplifier *IS* either evil or a moron. Smurf amplifier, 0WN3D unix box, open mail relay, who cares. LART 'em till they glow then shoot 'em in the dark.(Not that I'm not sympathetic to the difficulties of being a small ISP: I just don't think there's any excuse for operating an abusable data amplifier.)
The power needed to drive that bus will be 1/2 * N * f * C * V^2 = 0.5 * 256 * 2.5*10^9 * 10*10^-12 * 3.3^2 = 35Watts.
35 watts just for interconnect! Even if the optical interfaces consume a whopping 250mW each, you can still afford 140 of them for the power cost of copper.
But wait, there's more: modern CPUs need a huge L2 cache to compensate for the narrow pipe to main RAM. If you widen the pipe, you can get away with a lot less L2 cache, which saves a lot of power (cache is typically static RAM, which has 4 to 6 transistors per bit and sucks a lot of power). Optical interconnect can potentially provide a dedicated link between each RAM unit and the CPU. The latency will probably be higher, which will penalize things like office suites that have a random pattern of memory accesses, but signal processing, graphics, and technical calculations will be blazingly fast.
If there are several cables that are acceptable, you then ask "Which is cheapest?" and "Which is most durable?" and the answer for those is a resounding "plastic". Frankly, using glass fiber to move an audio signal a couple of meters would be really, really dumb.
Even with support from the USSR, the US was winning the war at the time of withdrawal. The withdrawal occurred because of the stunning PR incompetence of the US government. They didn't understand the tremendous power of an extended publicity campaign. They could probably have gotten support for a proxy war against the USSR, but they were silent and all that the public saw was an endless parade of body bags, year after year after year. As there was no strong leadership, the American public grew tired and ended the campaign.
Yet another major campaign in the Cold War, again a proxy war in fairly worthless territory, territory that neither nation would have bothered with were it not for the other superpower. Again, the Soviets sent vast amounts of materiel into the theatre, and again the US-supported forces destroyed most of what they sent in. Unlike the Vietnam War, the Soviets also sent lots of soldiers into Afghanistan, which was a lethal US-funded meat grinder.Something you have to understand about the Soviets was that their technology was not efficient. Compare to US factories, it was much more expensive for them to build a tank or fighter. The effectiveness and quality of Soviet war machines also tended to be rather low. The net result is that it cost the Soviets many more man-hours to field a credible military force. So when they sent in a tank that got promptly bombed by US-supported forces, they had to divert a lot more industrial capacity away from luxury goods, research and development, and so forth. At the same time, they spent far to much of the remaining industrial and R&D capacity trying to outdo the Strategic Defense Initiative. It's also worth pointing out the substantial diversion of Soviet R&D during the Vietnam Proxy War as they tried to compete with the US Apollo project.
Put all this together: previous costly war with no obvious victory and simultaneous loss at a technical competition, currently costly war with no obvious victory and heavy personnel losses and an even bigger unwinnable technological competition. It broke the will of the Soviet government and impoverished the people. The loss in Afghanistan was the straw that broke their back.
Only with extensive support from a superpower, and that ain't gonna happen for Al Quaida or the Taliban. Especially since the US-aligned Special Operations groups will be using what are, frankly, guerrilla tactics. It's going to be guerrilla versus guerrilla, only the US guerrillas will have C-130s full of materiel arriving as needed, good air support, night-vision scopes, satellite reconnaissance, encrypted spread-spectrum radios, and so forth.The US also has a major advantage: they are not trying to conquer and hold Afghanistan like the Soviets were. They are simply trying to kill and disrupt a certain few thousand people. Also, unlike Israel or Iran, the only US criterion for an Afghan govt is peacefulness, stability, and cooperation with US intelligence. They won't be trying to prop up a violent government to fight a proxy war against a major power (in fact, they'll be specifically avoiding such a govt).
Civilization is a continuous process, and a rather chaotic and poorly controlled one at that. There are human forces continually working for and against it, and occassionally the opposition wins out. The question is not how far can civilization rise, but how far can it fall. Historically, the most spectacular falls -- falls that brought the societies involved to their knees -- have occurred when the populace has become weak and a central imperial power has become strong. Think Imperial Rome shortly before the fall, Imperial Japan in the early 20th century where the overenthusiasm of the powerful military class brought the nation to ruin, Germany during the Third Reich years, or America during the War Between the States. Power disparities invariably breed violent revolution. Conversely, power equality and lack of social classes tends to breed stability, peace, and liberty.
That was my point. The government always has the sword. The only question is whether they will be able to push you around with it, or whether you can push back. It's not so much a threat as an insurance policy. Pushing around a citizen might be illegal, but pushing around an armed citizen is unhealthy. A paper justice system is easy to corrrupt, but no number of corrupt judges can sew the ears back on a dirty cop.