I'm disappointed that millimeter wave scanning and Z-backscatter hasn't yet made it to nightclubs. Security there can be more intrusive than at airports. Nightclub goons actually pat you down, which TSA doesn't do.
It would be fun to have the scans of people coming in on monitors around the club. Wny not? The clubbing crowd isn't that modest.
That's the trouble with doing this as an afterthought. The implementation is lame.
That's a terrible implementation.
There are several good ways to do it, most of which involve integration with the engine control system. One would be to provide a mechanical actuator which pushes back against the accelerator when the vehicle has reached the speed limit. A strong push can override it, but if you just press normally, you'll end up going exactly the speed limit. This is how Boeing does aircraft operating envelope protection in their fly-by-wire aircraft. It takes about 25 pounds of force to push through what the computers think the aircraft shouldn't be doing, and in a dire emergency, the pilot can choose to do that.
Some mid-1980s cars had something like that, but for a different reason. Computer
engine control was just becoming widespread, but the computers weren't yet trusted enough to be given full authority over the throttle. So there was a bell-crank arrangement, with the accelerator pedal and a powered actuator both linked to the throttle through springs. You could feel some pushback when the control system wanted to briefly back off the throttle while the transmission shifted. Few drivers even noticed.
This program has a sizable following in Japan; there's a whole "virtual idol" thing. The English versions just have voices, but the Japanese versions come with artwork and background information about the "singer". There are suspicions that some songs in anime were cranked out with this program.
Currently, the main limitation is that setting up the program for a new voice is a big job. There's a training process which requires the singer to sing some specific material to capture the needed phonemes. Once someone figures out a way to do that from existing recordings, it will be possible to emulate existing performers.
In other words, instant cover versions of classic songs.
That's going to give the RIAA headaches. Anybody can make a cover version of any song and pay just a statutory royalty. The RIAA likes that, because they get most of the money and the composers don't get much. But with this technology, the RIAA's function can be automated out of existence.
When this machine learns your job, what are you going to do?
Didn't anybody read the paper? It's not difficult, although it is somewhat chilling. It's intended to control weapon selection (chaingun, RPG-sized missile, or bunker-buster sized bomb), balancing military necessity (set by the operator) with collateral damage. This info is also used for selecting a firing position (get into a good position to use the chaingun, or just blow up the whole target area?)
Right now, we have weapons that are autonomous after launch, but dumb. This is, in a way, a step up.
You want to see good F/X? See "Angels and Demons". That wasn't filmed at the Vatican. The Vatican scenes, inside and out, were filmed in LA. It was done with partial sets, CG sets, green screen work, miniatures, matchmoves, and computer generated crowds. Can you tell?
Star Dreck was an easy F/X job. Anybody can do 3D spaceships. It's faking the commonplace that's tough.
We need to automate the generation and production of music, and crush the music industry like a bug.
Listen to this sample. That was created with Yamaha Vocaloid. The product sells for $179.95.
It's better than many singers. We're getting close.
This technology is like MIDI players, a generation later. You need the composition and instrument models. Then the player puts it all together. You can mix and match; choose a different singer or instruments. (Question: is there enough compute power in an iPhone to run this?)
If this catches on, the music industry will be crushed.
There's still a need for composers. Easy Music Composer isn't quite good enough. Yet.
It's pretty much a given that any facility that a world leader will spend a lot of time in, will have a safe room (bunker, if you will). I'd be fairly confident that the Whitehouse has one. The Pentagon is one.
The White House has a bunker, but it's not deep enough to survive a precision nuclear strike. There was a plan during the Ford administration to build a serious bunker and tunnel system connecting the White House and the Pentagon, about a mile down, but it probably wasn't built.
The White House has a tunnel to the Treasury building, and that's well known.
RISC. RISC allowed building simple CPUs that executed one instruction per clock. But once superscalar technology was developed, with more than one instruction per clock, RISC had to keep up. RISC CPUs became as complex as CISC CPUs, and the code density was worse. In the end, RISC was a lose, except at the very low end, like Atmel microcontrollers.
E-beam IC lithography. Exposing an IC with an electron beam, rather than "light" (which is now coming up on the soft X-ray end of the spectrum) has been a promising technology since the 1970s. No mask is required; just active steering of the electron beam by a computer. It works just fine. Line widths are better than what can be achieved with light and masks. It's just too slow.
Solid state magnetic memory. There have been many schemes for magnetic storage without moving parts. Core memory, of course. Magnetic bubbles. Ferroelectric RAM. All work technically, but have never had much market share.
Cryrogenic computing. This goes back to the early 1960s. NSA and IBM put a huge amount of effort into trying to make this work. They had gigahertz logic in the 1960s. The problem was that the gates could be made very fast, but not very small. IBM tried again with Josephson junctions. There's even a plan floating around DoD for a cyrogenic supercomputer. All this stuff works, but mainstream technology always ended up passing the technologies that ran in liquid helium.
Smoke printing. This is a forgotten idea. Write a charge pattern on the paper, run it through a smoke cloud of toner-like material, then fuse the toner. It's like laser printing, but without the photoconductive drum. The problem is that the process is very sensitive to humidity, and a printing technology that requires such tight environmental controls isn't worth the trouble when there are such good alternatives.
Shape-memory alloys. These were once touted as a new kind of motor, and a way to make robotic muscles. Run current through them, and they bend. The problem is that it takes a lot of current (because it's the heating that does it) and the actuators are slow.
Circuit-switched packet switching. It's quite possible to have useful circuit-switched data networks. Tymnet and Telenet, in the 1970s and 1980s, worked that way, as did X.25. At one point, this looked like the future, because congestion and quality of service can be better managed in a circuit-switched system. Telcos like this kind of thing, because it leads to connection-oriented billing. But pure datagrams won out, mainly because bulk bandwidth became cheap enough that the middle of the network could run at low load factors.
Wireless power transmission Not just Tesla; remember "powersats" and "rectennas"? A Japanese project once tried microwave power transmission between two islands. It worked, but wasn't efficient enough to be useful. We may see a comeback of this in the form of short-range wireless charging systems.
Very Long Instruction Word machines. Each word contains multiple instructions, executed simultaneously. The Itanium is an example of this class of architecture. The problem is that the compiler has to be very, very smart to code all the concurrency into the instructions. There doesn't seem to be a performance gain over more classical architectures. This is the curse of unusual architectures; MIMD machines, dataflow machines, hypercubes, perfect shuffle machines, and similar exotic ideas have come and gone. These machines can and have been built, but are very hard to program.
Wrist-mounted devices From Dick Tracy's two-way wrist radio to the HP-01, no wrist-mounted gadget with much more functionality than a watch has ever caught on. Around 1998, there was a flood of wrist pagers; that died out quickly. Even though one could cram considerable functionality into a watch-sized device today, there's little interest in doing so.
The Lisa was a useful little machine. The 128K Mac wasn't. In fact, the 128K Mac was an abysmal flop.
The Jobs Reality Distortion Field tends to rewrite history here. Because Jobs wasn't behind the Lisa. He wanted it to fail.
The Lisa had enough resources to be useful - 1MB of RAM, a hard drive, and an MMU. Unfortunately, 1MB of RAM cost too much in 1983, Apple's hard drive didn't work very well, and because of a design problem with instruction backout in the M68000, the MMU (not a chip, one built out of smaller scale ICs) couldn't really do paging right; the compiler had to avoid certain instructions.
In comparison, the 128K Mac was a joke. The OS and all your files had to fit on one floppy. The machine only had one built-in floppy drive. Sales were very low for the first few years. The Mac was a failure until hardware got cheap enough that its specs could be built up to Lisa levels. The Lisa was just a few years too early.
I built up the circuit as a SPICE model, and while it amplifies, it doesn't filter much. That weird filtering circuit in the emitter leg doesn't seem to accomplish anything. Treating the piezo microphone as a voltage source with a 1K resistance, generating a 1KHz input signal at 0.005V (based on a Murata piezo buzzer data sheet), what comes out is a voltage swing of about 0.6V at 1KHz, with a DC offset of 2.8V. The filtering seems to be insensitive to RM; changing RM from 10 ohms to 10 megohms doesn't do much to the output waveform. The 100K pot was adjusted until the voltage across RE was 3.3V, as specified. (This happens with the top end of the pot at 4.4K).
Why didn't they just put a nice simple low-pass filter on the output, instead of trying to get cute and put it in the emitter lead? And shouldn't there be a diode in there somewhere, to extract the waveform's envelope?
I actually built something like this in my teenage years, and had it hooked up to a surplus chart recorder (mirror galvanometer, phototube, relays, and motors, a mechanized Wheatstone bridge). (This dates me.) Mine worked.
Something like Second Life 3.x, or the virtual world in Snow Crash. Or GTA as a MMORPG.
(I'm surprised that there isn't an online version of GTA yet. Admittedly it's
tough to do well until the lag problem is solved. We need networks where you're guaranteed
about 10KB/s with under 50ms of round trip delay, for the data that really has to be
timely. The rest of the data (geometry updates, etc.) can have far more lag, but a
fraction of the data needs priority. The QoS people need to get their act together, so that clients and servers can request a low-bandwidth low-latency end to end path. To make this work, the bandwidth has to be limited.)
Yes. These are MIT students, remember. Now, if they designed something simple that got the same results as a properly calibrated A-law sound level meter, that would be useful. Or, for example, they could use the microprocessor to do an integrating dosimeter calculation, so you know when you've overdosed on live music. That would be useful to do cheaply, because noise dosimeters are still expensive, over $1000.
And, typically, we have a link to an article that plagiarized the entire story from the original in Women's Wear Daily. (Of course, being WWD, they go on to mention what the women on the panel were wearing.)
The real revolution hasn't been driven by the Internet. It's been driven by cheap disk space. Now everybody can store multiple copies of incredible amounts of dreck. If disk space cost like it did ten years ago, but we had today's bandwidth, there'd be far less junk around.
The movie industry has been losing the on-line battle for decades. When videotape first got cheap enough to rent, the movie industry thought a fair rental price should be equal to four theater tickets. Plus popcorn. Really.
Unlike the RIAA, the movie industry has failed at retail price maintenance. Music CDs are still selling around $17, while movie CDs tend to run less, and gradually move to the $4.99 bargain bin.
And that's without the Internet.
because it turns out that there were quite a number of spies in the U.S. government and other key institutions of American society.
True. But they were mostly useless. One of the more amusing Soviet documents to surface from that era was a grumbling memo from KGB Moscow Central. They were complaining that too many useless agents had been recruited in places like the State Department, and not enough in atomic and other defense programs.
Some well-known items:
The Rosenbergs really were guilty, both of them, but they weren't most useful spies. The Venona transcripts are fascinating, because you can see why the US was so frantic at the time. The intercepts made it clear that the Russians had spies inside the nuclear program, but didn't identify them. It took years of messages (stuff like "met ANTENNA in Baltimore") before enough info was collected to identify the leak.
There are some later interesting disclosures about how the Soviet Union financed the Vietnam-era anti-war movement in the US.
For an good background on that era from the Soviet side, read Anatoly Dobrynin's memoirs. He was Soviet ambassador to the US from 1962 to 1986, and after the Cold War wound down, wrote it all up. Dobrynin became an ambassador due to a whim of Stalin's. One day, Stalin was frustrated with his diplomats, who were mostly old guys left over from the Revolution. He said something like "We need new Soviet men in this job, like young aircraft designers." The next day, Anatoly Dobrynin, young aircraft designer, was taken from his drawing board (literally) by KGB agents and shipped to Moscow, to attend the Higher Diplomatic Academy. And no, he wasn't told why at the time.
This could be useful. It will result in an official DoD list of known spammers. That will make prosecutions easier. And the "attack on Government computer" provision in the Computer Crime Act will apply.
If someone from DISA pushes hard enough, the FBI can be tasked to take down the top spammers. It doesn't matter where they are; if the U.S. Government is annoyed enough with them, they can be shut down. That's what the State Department is for.
If one spammer a month went to jail, there would be a huge drop in spam. We see a big drop each time a big spammer goes down. There just aren't that many players.
Sad thing here, 20yrs ago I could pick up just about anything I needed for my electronics needs.
But now we have Digi-Key. Parts ordering is better than ever. Did you ever order from Allied Radio? It took weeks, and about 5-10% of the parts would be out of stock. Hamilton/Avnet wouldn't even take orders from individuals. Don't complain.
There's so much more information available on line. Online PCB design and fab works very well and isn't that expensive. Free CAD tools are available. Even SPICE simulation is available for free, and it works quite well. The big headache with do-it-yourself electronics today is that surface mount assembly is beyond the ability of most hobbyists, and the newer parts are surface mount only.
The real problem is that it's tough for a kid to build anything cool out of parts. Forty years ago, it was cool to build an intercom, or a touch lamp. Today, not. Consumer electronics is way ahead of DIY electronics.
It's a neat problem. Some years back I was almost sucked into There, Inc. to work on that specific problem.
There are two issues; implementation scaling and game mechanics.
Second Life is one big world, but there are severe limits on how many people can be in the same area at a time. They really haven't solved their scaling problem. This is a tough design problem. But it's not unsolveable. I was at one time looking at an architecture where the world is divided into hexagons, with a moderate overlap between adjacent hexagons. Within the overlap area, servers negotiate with the server for the neighboring hexes (never more than two; that's the advantage of going hexagonal) over who's in charge of characters and items in the overlap area. Overloaded hexagons are subdivided, so more servers can be brought to bear on heavily loaded hexagons. Flying over the world creates problems, but by careful use of level of detail problem, and interposing fog and clouds in difficult situations, that could be handled. I think this is solveable today. There's going to be a lot of gigabit Ethernet cable in the server farm.
The gameplay problem is that everyone may want to go to the hot spots. Some games have more problems with this than others. Star Wars, big problem. GTA, not so much.
Terrible article. See Robots Dreams for what's really going in Japanese hobbyist robotics. Especially what's happening with small humanoid robot competitions. Obstacle courses are routine now.
This stuff is way ahead of the Lego Mindstorms, Battlebots, anf FIRST stuff you see in the US.
I'd like to see Big Dog balance technology scaled down to toy size. It's not inherently expensive. All the cleverness is in the software and the math.
I've never been impressed with the mania for "zero copy" systems. On modern CPUs, copying of data handled very recently is cheap, because it's already in the faster caches. On the other hand, mucking about with the MMU to move pages from one address space to another tends to be expensive, especially if cache flushing is required. Mach made that design mistake.
I'm old enough to remember when the "sockets" API was developed. We'd been using a very early 3COM TCP/IP package, "UNET", which predated BSD networking. It simply used "open", "read", and "write", rather than special "socket" calls. Adding extra calls was very Berkeley; they were writing alongside the UNIX kernel, not fully integrating their own stuff. There was no reason not to have "read" and "write" work on sockets, and in some operating systems, they do.
Bear in mind that BSD didn't have threads. Hence the need for the polled "select" model.
If you want to see interprocess communication done right, look at QNX. Their "MsgSend", "MsgReceive", and "MsgReply" model allows one program to call another. If you want networking to call the application, that's the way to do it. It's a proven model, and it's fast enough that I've pumped uncompressed video through it with message passing using about 3% of a Pentium III class CPU.
By the way, bear in mind that ACM Queue is just a sort of blog. That's not a refereed paper.
I think there was at least one case on this involving hotels. Back in the early VHS era, many hotels provided VHS players for guests, and lent out tapes at the front desk. No problem there, at least in the US; that's just the "first sale" doctrine.
Better hotels would deliver tapes via room service. This was labor-intensive. Some hotel then realized that it would easier to centralize all the VCRs, and just have someone in an office put the requested tape in the VCR when requested. This was the beginning of "video on demand".
That was held not to be a copyright infringement, even though the hotel was in a sense "distributing" the content.
Now, of course, there are "video on demand" systems for hotels. But they usually have contractual relationships with all their sources; they're not just buying VHS tapes at retail.
Slashdot Mirror
I'm disappointed that millimeter wave scanning and Z-backscatter hasn't yet made it to nightclubs. Security there can be more intrusive than at airports. Nightclub goons actually pat you down, which TSA doesn't do.
It would be fun to have the scans of people coming in on monitors around the club. Wny not? The clubbing crowd isn't that modest.
That's the trouble with doing this as an afterthought. The implementation is lame.
That's a terrible implementation.
There are several good ways to do it, most of which involve integration with the engine control system. One would be to provide a mechanical actuator which pushes back against the accelerator when the vehicle has reached the speed limit. A strong push can override it, but if you just press normally, you'll end up going exactly the speed limit. This is how Boeing does aircraft operating envelope protection in their fly-by-wire aircraft. It takes about 25 pounds of force to push through what the computers think the aircraft shouldn't be doing, and in a dire emergency, the pilot can choose to do that.
Some mid-1980s cars had something like that, but for a different reason. Computer engine control was just becoming widespread, but the computers weren't yet trusted enough to be given full authority over the throttle. So there was a bell-crank arrangement, with the accelerator pedal and a powered actuator both linked to the throttle through springs. You could feel some pushback when the control system wanted to briefly back off the throttle while the transmission shifted. Few drivers even noticed.
Killing the engine is just lame.
That was Version 1 of Vocaloid. Listen to Japanese pop generated with Version 2 (MP3). Version 2 sounds better than Version 1; they're getting to lead-singer quality.
This program has a sizable following in Japan; there's a whole "virtual idol" thing. The English versions just have voices, but the Japanese versions come with artwork and background information about the "singer". There are suspicions that some songs in anime were cranked out with this program.
Currently, the main limitation is that setting up the program for a new voice is a big job. There's a training process which requires the singer to sing some specific material to capture the needed phonemes. Once someone figures out a way to do that from existing recordings, it will be possible to emulate existing performers.
In other words, instant cover versions of classic songs. That's going to give the RIAA headaches. Anybody can make a cover version of any song and pay just a statutory royalty. The RIAA likes that, because they get most of the money and the composers don't get much. But with this technology, the RIAA's function can be automated out of existence.
When this machine learns your job, what are you going to do?
Didn't anybody read the paper? It's not difficult, although it is somewhat chilling. It's intended to control weapon selection (chaingun, RPG-sized missile, or bunker-buster sized bomb), balancing military necessity (set by the operator) with collateral damage. This info is also used for selecting a firing position (get into a good position to use the chaingun, or just blow up the whole target area?)
Right now, we have weapons that are autonomous after launch, but dumb. This is, in a way, a step up.
Next, Google autonomous aircraft. Big ones for overall views, little ones for street views. Small boats for waterways and coastlines.
Then, Google Humans. Face pictures of everybody on the planet.
Laugh now. Someday they'll be in charge.
You want to see good F/X? See "Angels and Demons". That wasn't filmed at the Vatican. The Vatican scenes, inside and out, were filmed in LA. It was done with partial sets, CG sets, green screen work, miniatures, matchmoves, and computer generated crowds. Can you tell?
Star Dreck was an easy F/X job. Anybody can do 3D spaceships. It's faking the commonplace that's tough.
We need to automate the generation and production of music, and crush the music industry like a bug.
Listen to this sample. That was created with Yamaha Vocaloid. The product sells for $179.95. It's better than many singers. We're getting close.
This technology is like MIDI players, a generation later. You need the composition and instrument models. Then the player puts it all together. You can mix and match; choose a different singer or instruments. (Question: is there enough compute power in an iPhone to run this?)
If this catches on, the music industry will be crushed.
There's still a need for composers. Easy Music Composer isn't quite good enough. Yet.
It's pretty much a given that any facility that a world leader will spend a lot of time in, will have a safe room (bunker, if you will). I'd be fairly confident that the Whitehouse has one. The Pentagon is one.
The White House has a bunker, but it's not deep enough to survive a precision nuclear strike. There was a plan during the Ford administration to build a serious bunker and tunnel system connecting the White House and the Pentagon, about a mile down, but it probably wasn't built. The White House has a tunnel to the Treasury building, and that's well known.
The USSR, though, really did build secret underground tunnels and subways between the Kremlin and distant bunkers.
Not products, technologies.
The Lisa was a useful little machine. The 128K Mac wasn't. In fact, the 128K Mac was an abysmal flop.
The Jobs Reality Distortion Field tends to rewrite history here. Because Jobs wasn't behind the Lisa. He wanted it to fail.
The Lisa had enough resources to be useful - 1MB of RAM, a hard drive, and an MMU. Unfortunately, 1MB of RAM cost too much in 1983, Apple's hard drive didn't work very well, and because of a design problem with instruction backout in the M68000, the MMU (not a chip, one built out of smaller scale ICs) couldn't really do paging right; the compiler had to avoid certain instructions.
In comparison, the 128K Mac was a joke. The OS and all your files had to fit on one floppy. The machine only had one built-in floppy drive. Sales were very low for the first few years. The Mac was a failure until hardware got cheap enough that its specs could be built up to Lisa levels. The Lisa was just a few years too early.
I built up the circuit as a SPICE model, and while it amplifies, it doesn't filter much. That weird filtering circuit in the emitter leg doesn't seem to accomplish anything. Treating the piezo microphone as a voltage source with a 1K resistance, generating a 1KHz input signal at 0.005V (based on a Murata piezo buzzer data sheet), what comes out is a voltage swing of about 0.6V at 1KHz, with a DC offset of 2.8V. The filtering seems to be insensitive to RM; changing RM from 10 ohms to 10 megohms doesn't do much to the output waveform. The 100K pot was adjusted until the voltage across RE was 3.3V, as specified. (This happens with the top end of the pot at 4.4K).
Why didn't they just put a nice simple low-pass filter on the output, instead of trying to get cute and put it in the emitter lead? And shouldn't there be a diode in there somewhere, to extract the waveform's envelope?
I actually built something like this in my teenage years, and had it hooked up to a surplus chart recorder (mirror galvanometer, phototube, relays, and motors, a mechanized Wheatstone bridge). (This dates me.) Mine worked.
Something like Second Life 3.x, or the virtual world in Snow Crash. Or GTA as a MMORPG.
(I'm surprised that there isn't an online version of GTA yet. Admittedly it's tough to do well until the lag problem is solved. We need networks where you're guaranteed about 10KB/s with under 50ms of round trip delay, for the data that really has to be timely. The rest of the data (geometry updates, etc.) can have far more lag, but a fraction of the data needs priority. The QoS people need to get their act together, so that clients and servers can request a low-bandwidth low-latency end to end path. To make this work, the bandwidth has to be limited.)
Yes. These are MIT students, remember. Now, if they designed something simple that got the same results as a properly calibrated A-law sound level meter, that would be useful. Or, for example, they could use the microprocessor to do an integrating dosimeter calculation, so you know when you've overdosed on live music. That would be useful to do cheaply, because noise dosimeters are still expensive, over $1000.
And, typically, we have a link to an article that plagiarized the entire story from the original in Women's Wear Daily. (Of course, being WWD, they go on to mention what the women on the panel were wearing.)
The real revolution hasn't been driven by the Internet. It's been driven by cheap disk space. Now everybody can store multiple copies of incredible amounts of dreck. If disk space cost like it did ten years ago, but we had today's bandwidth, there'd be far less junk around.
The movie industry has been losing the on-line battle for decades. When videotape first got cheap enough to rent, the movie industry thought a fair rental price should be equal to four theater tickets. Plus popcorn. Really.
Unlike the RIAA, the movie industry has failed at retail price maintenance. Music CDs are still selling around $17, while movie CDs tend to run less, and gradually move to the $4.99 bargain bin. And that's without the Internet.
I want my city to have a Steel Pipe Snow Melting System. And underground freight tunnels. And a Continuous Transit System with Sub-Surface Moving Platforms. And rooftop heliports. And skyscraper restaurants.
Put it on Wikileaks. Send it anonymously to the New York Times, the Washington Post, and a few key congressional offices. That's a good start.
When in doubt, ask what would Herblock say?
because it turns out that there were quite a number of spies in the U.S. government and other key institutions of American society.
True. But they were mostly useless. One of the more amusing Soviet documents to surface from that era was a grumbling memo from KGB Moscow Central. They were complaining that too many useless agents had been recruited in places like the State Department, and not enough in atomic and other defense programs.
Some well-known items:
For an good background on that era from the Soviet side, read Anatoly Dobrynin's memoirs. He was Soviet ambassador to the US from 1962 to 1986, and after the Cold War wound down, wrote it all up. Dobrynin became an ambassador due to a whim of Stalin's. One day, Stalin was frustrated with his diplomats, who were mostly old guys left over from the Revolution. He said something like "We need new Soviet men in this job, like young aircraft designers." The next day, Anatoly Dobrynin, young aircraft designer, was taken from his drawing board (literally) by KGB agents and shipped to Moscow, to attend the Higher Diplomatic Academy. And no, he wasn't told why at the time.
This could be useful. It will result in an official DoD list of known spammers. That will make prosecutions easier. And the "attack on Government computer" provision in the Computer Crime Act will apply.
If someone from DISA pushes hard enough, the FBI can be tasked to take down the top spammers. It doesn't matter where they are; if the U.S. Government is annoyed enough with them, they can be shut down. That's what the State Department is for.
If one spammer a month went to jail, there would be a huge drop in spam. We see a big drop each time a big spammer goes down. There just aren't that many players.
This should have been done thirty years ago.
Sad thing here, 20yrs ago I could pick up just about anything I needed for my electronics needs.
But now we have Digi-Key. Parts ordering is better than ever. Did you ever order from Allied Radio? It took weeks, and about 5-10% of the parts would be out of stock. Hamilton/Avnet wouldn't even take orders from individuals. Don't complain.
There's so much more information available on line. Online PCB design and fab works very well and isn't that expensive. Free CAD tools are available. Even SPICE simulation is available for free, and it works quite well. The big headache with do-it-yourself electronics today is that surface mount assembly is beyond the ability of most hobbyists, and the newer parts are surface mount only.
The real problem is that it's tough for a kid to build anything cool out of parts. Forty years ago, it was cool to build an intercom, or a touch lamp. Today, not. Consumer electronics is way ahead of DIY electronics.
It's a neat problem. Some years back I was almost sucked into There, Inc. to work on that specific problem.
There are two issues; implementation scaling and game mechanics.
Second Life is one big world, but there are severe limits on how many people can be in the same area at a time. They really haven't solved their scaling problem. This is a tough design problem. But it's not unsolveable. I was at one time looking at an architecture where the world is divided into hexagons, with a moderate overlap between adjacent hexagons. Within the overlap area, servers negotiate with the server for the neighboring hexes (never more than two; that's the advantage of going hexagonal) over who's in charge of characters and items in the overlap area. Overloaded hexagons are subdivided, so more servers can be brought to bear on heavily loaded hexagons. Flying over the world creates problems, but by careful use of level of detail problem, and interposing fog and clouds in difficult situations, that could be handled. I think this is solveable today. There's going to be a lot of gigabit Ethernet cable in the server farm.
The gameplay problem is that everyone may want to go to the hot spots. Some games have more problems with this than others. Star Wars, big problem. GTA, not so much.
Terrible article. See Robots Dreams for what's really going in Japanese hobbyist robotics. Especially what's happening with small humanoid robot competitions. Obstacle courses are routine now.
This stuff is way ahead of the Lego Mindstorms, Battlebots, anf FIRST stuff you see in the US.
I'd like to see Big Dog balance technology scaled down to toy size. It's not inherently expensive. All the cleverness is in the software and the math.
I've never been impressed with the mania for "zero copy" systems. On modern CPUs, copying of data handled very recently is cheap, because it's already in the faster caches. On the other hand, mucking about with the MMU to move pages from one address space to another tends to be expensive, especially if cache flushing is required. Mach made that design mistake.
I'm old enough to remember when the "sockets" API was developed. We'd been using a very early 3COM TCP/IP package, "UNET", which predated BSD networking. It simply used "open", "read", and "write", rather than special "socket" calls. Adding extra calls was very Berkeley; they were writing alongside the UNIX kernel, not fully integrating their own stuff. There was no reason not to have "read" and "write" work on sockets, and in some operating systems, they do.
Bear in mind that BSD didn't have threads. Hence the need for the polled "select" model.
If you want to see interprocess communication done right, look at QNX. Their "MsgSend", "MsgReceive", and "MsgReply" model allows one program to call another. If you want networking to call the application, that's the way to do it. It's a proven model, and it's fast enough that I've pumped uncompressed video through it with message passing using about 3% of a Pentium III class CPU.
By the way, bear in mind that ACM Queue is just a sort of blog. That's not a refereed paper.
Ask if the "social networking" restrictions apply while playing golf.
I think there was at least one case on this involving hotels. Back in the early VHS era, many hotels provided VHS players for guests, and lent out tapes at the front desk. No problem there, at least in the US; that's just the "first sale" doctrine.
Better hotels would deliver tapes via room service. This was labor-intensive. Some hotel then realized that it would easier to centralize all the VCRs, and just have someone in an office put the requested tape in the VCR when requested. This was the beginning of "video on demand".
That was held not to be a copyright infringement, even though the hotel was in a sense "distributing" the content.
Now, of course, there are "video on demand" systems for hotels. But they usually have contractual relationships with all their sources; they're not just buying VHS tapes at retail.