RF Monolithics make a bunch of one-chip radio solutions. OK, so it requires some electronics to interface to the PC, but not much. Take a look at the TX5000 and RX5000 datasheets. They're low power, too, which is nice for battery powered applications.
If you want to connect multiple devices on the one frequency, then you have to have a transmitter and receiver at each unit, and have the master run a polling loop. So that all gets a bit tricky.
Another alternative is Bluetooth. Magical, but still forthcoming, and still very expensive at the moment.
like most Australian inventions, this one has only found serious commercial backing overseas
I had the brain drain thing driven home to me recently. I went to a "smart kids" program in high school, here in Western Australia. Last weeekend, my school had an "old students day" (preparatory to the government demolishing the school at the end of the year -- nice one, Richard). I went there looking for my class mates. Mostly, the few that I found were the ones that, like myself, had returned from working overseas or in Australia's eastern states. The rest were still over there.
That's right, way more than half my "smart kids" class had to leave Western Australia to find challenging and/or financially rewarding work.
Has anyone tried writing a complete virtual processor/virtual peripheral system that performs dynamic binary translation between instruction sets?
There are a few movements in this direction, that emulate the processor, but not all the other hardware you'd expect to find in a traditional x86 PC.
One such is HP's Dynamo. It emulates an HP PA-8000 instruction set on, get this, an HP PA-8000. This emulation is faster than the native code, because it can do better optimisation.
Sun's MAJC runs a similar optimising emulator, but it emulates a Java machine.
I would guess that nobody's tried to emulate a complete x86 PC on a non-x86 platform largely because they'd also have to emulate all the features other than the processor: mouse, keyboard, serial, parallel, IDE, USB, etc.
If you look at http://telstra.com.au/bigpond/direct/aboutnet.htm, you'll see that Telstra's fattest link to the US is in fact from Perth. So given that you're using Telstra, that's quite a reasonable routing. Presumably, Telstra not being part of this alliance, they're not yet using any of this fat new pipe.
And, hey, routing can be pretty f***ed up. I've sent packets from iPrimus in Perth to iinet in Perth, and had them go via Adelaide, Melbourne and Sydney. (For US readers, that's about 5000 miles for a 5 mile straight-line journey.)
Here in Australia (and, I suspect, in the Rest Of The World), the default calling plan works like this: whoever originates the call, pays for it.
Now, Rest Of The World readers are probably thinking, "Well, duh. But in the U.S., the default works like this: the owner of the cell phone pays for it.
This is, IMAO, one of the reasons cell phones are less widespread in the U.S. that they are in Rest Of The World. See if you can get a RotW-style calling plan; they're good, 'cos you don't have to fear people calling you up and costing you lots of money.
X10 is a protocol for controlling appliances via power lines. For example, if your computer was fitted with an X10 interface, you could use it to control any X10-aware devices on the same electrical system; including lightbulbs, toasters, TV's, air conditioners, whatever. Of course, you have to go to the grief of making those devices X10 aware; and usually also installing an isolator so that you can't control your neighbours lights, and he can't control yours.
I worked for a company in Australia, which manufactured laser tag equipment and exported it all over the world. The US sister company needed experience, and the logical place to go to find that was the Australian company. So the Australian company duly sold me down the river to go and work for the American company. I made US$36K while working in America, which ain't exactly slave wages.
But it could easily have been slave conditions. My H1B visa allowed me to work for this one specific company in the U.S., and no other. If they fired me, I had to leave the country within ten days. Ten days! Sell my car, arrange to move all my possessions internationally, deal with ending my rental lease, blah blah f***ing blah, all within ten days!
And eventually, this is what did happen. The company, as high-tech startups are wont to do, went belly-up, leaving me stranded. I applied to get a tourist visa. This is trivially easy to do from outside the country; but impossible from inside, due to the beauracracy. Basically, this is one of the many parts of the INS which has ground to a complete halt. While this was going on, I applied for a job with the local library system. It was a job they'd advertised a bunch of times before, but not been able to fill, as they hadn't been able to find anyone with the right skillset. I, however, matched perfectly. This is what H1B's are about -- no American qualified to fill the job, so give it to a foreigner. They were all set to pay me US$45K to US$61K, when we found out how hard the immgration stuff was going to be. It would take 3 to 4 months to process my visa, and this processing had to be completed before I could do any work at all for my new prospective employer. Understandably, they turned down the opportunity to hire me.
And that 3 to 4 months was at the good end of the year! A month later, visas ran out, meaning that there would be a six month wait for the new year (with a new visa allotment) to start, and then 3 to 4 months on top of that! Outside the Silicon Valley firms which are specifically geared up to import tech workers, what organisation has any idea of what its high-tech employee requirements are going to be three to ten months ahead?
There are more visas for this coming "year" (which starts in October), which means that there won't be the six month wait, but with the same number of beauracrats to process way more visas, guess what? That's right, that 3 to 4 month wait is going to blow way out.
So I had to leave the US, leaving behind all my friends. I should mention that my experience was in the area aerved by the California branch of the INS. It is the one completely snowed by the influx to Silicon Valley; and I understand that other regional INS's are more on top of things.
It is my opinion that these massive delays have been allowed to creep in by three things: first, of course, the rapid increase in influx of workers; second in the US immigration law's attitude that America is the greatest country in the world and that if we didn't have these laws everybody in the world, all six bilion of them, would be on our doorstep tomorrow; and third in the fact that the people the INS serves generally aren't voters.
With 2K (instruction, 14 bits per) linear code address space, [...] it ain't gonna run linux.
That's the limiting factor. You're right; obviously, you couldn't fit Linux in 2K words of code. So you have to expand the available code space. This is of course impossible -- the PIC is only capable of executing instructions from its internal store.
But it's possible to hang external memory off a PIC, whether it be a Dallas iButton or a more traditional RAM or ROM chip; and if you tried, you could probably build an interpreter to fit in the PICs internal instruction space, to execute "instructions" stored in this off-chip memory.
So yes, you could build a machine that ran Linux using a PIC as the processor; but Linux wouldn't be actually on board the PIC itself. Still, there are some nice small SMT EEPROM chips around, and you could fit one of those plus an SMT PIC on a pretty damn small board...
Incidentally, some of the newer PICs (e.g. the PIC16C77, I think) are not OTP; they're flash. Sweet.
I am totally with you. Few PCs have LS120 SuperDisk drives, and not many more have Zip drives. But CD ROM drives are ubiquitous. Carry a single CD with you, drop it in any PC anywhere, and bam, you're running Linux.
In the simple version, it would boot off CD, and look for local fd drives and hd partitions to auto-mount. The all-singing all-dancing version would scan those partitions, and look for a Windows registry with clues as to what name and IP number it should be; and wander out on the LAN and look for SMB shares to mount. It could come with tools to roll a new CD, with new apps etc.
Maybe it should be called Ronix, for Read-Only Linux.
Does anybody know if anything like this already exists?
It's because OpenBSD is the security-conscious version of BSD. They focus on having really good, really thorough security first, everything else second. As such, encrypting swap is a good, clever idea.
To connect two NICs (usually two PC's), you need a crossover cable. You may be able to find a picture of one in a manual for an Ethernet hub, particularly an older one.
4 & 5 and 7 & 8 are pairs too. They're not used by Cat5, 100Mbit; but they are used by Cat5e, 1Gbit; so hook 'em up now to save trouble later.
Now, 1 & 2 are the "transmit" pair; think of one line as the signal, and the other as the ground return for that signal. Each line needs its own return in order to get the groovy benefits of the magic of twisted pair. Similarly, 3 & 6 are the "receive" pair. When you hook a PC to a hub (the usual state of affairs), the "receive" and "transmit" are the other way 'round, so the PC transmits to the hub's receive, and vice versa, so all is happiness.
But when you're hooking two PC's together direct, if you used a straight cable, you'd be hooking one PC's transmit to the other PC's transmit, and the same for receive. No workee. So we swap those pairs:
So that's 1 -> 3, 2 -> 6, 3 -> 1, 6 -> 2. 4 & 5 and 7 & 8 are again not used, but hook 'em up anyway.
When you test your cable, you can buy a cheap cable checker that shows a little light for each line (try Weidmuller / Paladin Tools), or a multimeter, and this will tell you whether you have continuity on each line. However, to test the cable properly, you really need a much more expensive checker that tests if it's gonna work at 100Mbit. After all, at 100Mbit, each bit is only 3 metres long!
Yes, this is exactly what large chunks of Quarantine is about. The idea is that an experimental subject receives a neural mod that allows her to influence the result of a simple quantum event.
It's great as the premise for an SF novel (and it is a really good novel); but it's bogus in real life.
Something the article misses is that there is less incentive for people to write virii in the Linux space.
In the Windows space, if you want to see code you wrote running on thousands of machines, one of the quickest ways up is to write a virus. But in the Linux space, the quickest way up is to join an open-source project. Not only do you get to point to some tweaky little feature in KDE or whatever and say, "I wrote that!"; but you're also (generally) made to feel welcome and supported by your fellow project members.
The kids who download free music from a young age as a matter of course have little awareness that they are appropriating someone else's property.
Bullshit. They have plenty of awareness of what they're doing. A student in our school newspaper was quoted as saying something to the effect of, "Sure, everyone knows it's illegal, but they do it anyway because it's easy."
While they understand that it's illegal, I don't think they regard it as immoral. I used to be this way myself: way back when I owned a C-64, I used to pirate software because (a) I had no money, (b) everybody else was doing it, and (c) since I couldn't afford to buy the games anyway, I wasn't really taking money from their pocket, was I?
The immorality of removing someone's opportunity to make profit off their hard work is harder to see, particularly when everyone else is doing it, and particularly when you know that the original artists receive only a tiny share of the money anyhow).
I think that if music publishers made music available in a more direct way, and made a more direct pitch to the people doing the copying to please respect the rights of the artists (as opposed to jumping up and down and shouting, "Pirates are evil!"), that they might well find a satisfactory resolution.
Hmm... Try demo'ing it to them in a way that highlights the bad UI elements for what they are.
For example, if the page is bloated, demo it on a dial-up line, so it loads really slowly. Emphasise to them that this is the way 90% of their customers will experience it.
Have a user who is unfamiliar with the stupid-mouseover thing sitting in the driver's seat when you demo it. Tell him to walk in a particular direction which will highlight the stupidness of said "feature".
OK, so it isn't always easy to demo stuff in the way you want, but sometimes it's possible.
That's what this deal will get you. A whole bunch of different ISP's will be accessible on the cable that already runs to your door. Some, like RoadRunner, will charge you for access plus email, news, webspace, etc.; but some will offer a more minimalist service, and charge you less dollars accordingly.
And second, what kind of architecture - software and hardware -can do for this Internet generation what the Net protocols did for the last one?
Technology defines society. For example, it used to be perfectly acceptable to keep slaves, because farming and manufacturing were labour-intensive. It was only once it was economically feasible to do away with slavery that the bulk of public opinion shifted, because until then the people in power had an interest in it not shifting.
Then, too, society defines technology. The internet itself is a creation of the U.S. government's tendency to throw vast numbers of dollars at military projects. Same deal for modems.
So one of the questions Katz is trying to ask here is: What technology should we create, in order to shape society?
Jack William Bell (who wins the "Most Karma gained in a Day" prize) makes the point that this is far from the whole story, and that law and propaganda are extremely important. For example, you and I will always be able to get hold of a copy of DeCSS, regardless of what various laws may say, because we have a clue; but if it's illegal, then it can't be included in your facourite Linux distro, and thus will never be available to the bulk of the population. Freedom for a small elite is a pretty sucky kind of freedom.
But inroads can be made through a technological front. As Dr Caleb has mentioned, the Internet exists in places other than the US.
Here is one such technological front: a data haven. A bunch of gutsy servers located technically, physically or practically outside the jurisdictions that want to control the Internet. One such model is described in Neal Stephenson's excellent Cryptonomicon; another is being created by people building a free networking system that caches data anonymously, thus meaning that there's no-one to serve with an injunction (can somebody remember their URL?). Want DeCSS? Go to a data haven. (Want kiddie porn? Go to a data haven. Clearly, complete freedom has a bad side as well as a good side, but to an extent ya gotta accept both sides of the coin, or neither.)
And hey, is anyone other than me having trouble believing that Katz said "Yo"?
Your MS Money example is nice, but it's still computer-based, and therefore potentially tricky for Joe Average. Here's another analogy:
Imagine you have a piece of paper, with a coded message on it. You can easily copy the piece of paper (and the message), using a photocopier.
The point is: copying has nothing whatsoever to do with decryption.
Further, imagine that these pieces of paper with coded messages were your DVD collection. Your DVD player has to be able to decode the message to play the movie. If you want to be able to build your own DVD player ('cos the commercial ones suck), your player has to be able to decrypt the code, just like existing players.
Again, the point is: the ability to copy and the ability to decode are completely different. Decoding is necessary for playing; it's not necessary for copying.
We've been hearing recently about quantum cryptography, in which data is sent in an uncrackable format; and also about quantum computers, whose massively parallel computation ability might make current crypto trivially easy to break.
I know that people have had some success in the laboratory with both of these things; do you see them succeeding in the real world? How soon?
I'm fascinated that a number of hackers, with little in common other than a desire and an ability to fix a piece of the problem, can convene in an ad-hoc forum and fix the problem; probably before the problem's "owners" are even aware of its existence.
Is this the first time slashdot has been used in this way, as a real-time tool to coordinate a debugging effort?
Some of my friends at the University Computer Club of the University of Western Australia did a very similar thing. They had an amazingly antique terminal controller, which they had aquired in the same way they got most of their hardware: it was totally obselete, and therefore free.
But it didn't come with the keyboards; so they made keyboards themselves. Take three computer punch-cards, cut holes in the middle one, in exactly the fashion outlined on Jim's site, place aluminium foil "row" strips between the first and second card, and "column" strips between the second and third. Draw "keys" on the first card. Voila! Instant keyboard.
As the keyboard got older, it would get harder to use, and you'd have to bash the keys with a pen, and eventually throw it away and build a new one.
The fundamental problem is that US currency is so easy to copy. I have easily enough stuff lying around my office to produce reasonably realistic copies.
Software aimed at specifically recognising currency and stamps is foolish: it will only recognise certain kinds of currency and stamps; it won't reconise foreign stamps; and it won't recognise other paper instruments which we would rather not see forged (certificates, etc.)
Software aimed at making forgeries trackable is more thoughtful; but it has obvious privacy implications, and is potentially technically defeatable (as many readers have mentioned).
The fundamental solution is to make currency harder to forge. Australian currency notes, for example, are printed on a thin papery plastic instead of on paper; they have a piece of artwork partly printed on each side, so it is obvious if the artwork on the two sides is misaligned; and they have a transparent section, so it is obvious if it is printed on the wrong "paper". In a similar vein, the new US $20 note has a "color change" section that looks different when viewed from different angles.
Trying to fix the problem by limiting the technology in a thousand different scanners, printers and copiers is a bad approach: it's analogous to trying to cover for your lousy encryption by crippling everybody else's computer. The Right Thing is improve the technology in the money itself.
Slashdot Mirror
Once again, Intel have been beaten to the punch by their competitors: Transmeta recalled their chip a full 24 hours ago!
If you want to connect multiple devices on the one frequency, then you have to have a transmitter and receiver at each unit, and have the master run a polling loop. So that all gets a bit tricky.
Another alternative is Bluetooth. Magical, but still forthcoming, and still very expensive at the moment.
That's right, way more than half my "smart kids" class had to leave Western Australia to find challenging and/or financially rewarding work.
One such is HP's Dynamo. It emulates an HP PA-8000 instruction set on, get this, an HP PA-8000. This emulation is faster than the native code, because it can do better optimisation.
Sun's MAJC runs a similar optimising emulator, but it emulates a Java machine.
I would guess that nobody's tried to emulate a complete x86 PC on a non-x86 platform largely because they'd also have to emulate all the features other than the processor: mouse, keyboard, serial, parallel, IDE, USB, etc.
4 139.130.82.57 serial5-1-3.wel5.telstra.net
5 139.130.238.230 fddi0-0.wel-core2.perth.telstra.net
6 203.50.113.39 gigabitethernet4-0.wel-core3.perth.telstra.net
7 203.50.113.18 gigabitethernet4-0.wel-gw1.perth.telstra.net
8 203.50.126.30 pos1-0.paix1.paloalto.telstra.net
9 209.1.169.97 paix-f2-5.exodus.net
If you look at http://telstra.com.au/bigpond/direct/aboutnet.htm, you'll see that Telstra's fattest link to the US is in fact from Perth. So given that you're using Telstra, that's quite a reasonable routing. Presumably, Telstra not being part of this alliance, they're not yet using any of this fat new pipe.
And, hey, routing can be pretty f***ed up. I've sent packets from iPrimus in Perth to iinet in Perth, and had them go via Adelaide, Melbourne and Sydney. (For US readers, that's about 5000 miles for a 5 mile straight-line journey.)
Now, Rest Of The World readers are probably thinking, "Well, duh. But in the U.S., the default works like this: the owner of the cell phone pays for it.
This is, IMAO, one of the reasons cell phones are less widespread in the U.S. that they are in Rest Of The World. See if you can get a RotW-style calling plan; they're good, 'cos you don't have to fear people calling you up and costing you lots of money.
1. What about that face, huh? It's a government conspiracy! And I'm sure Microsoft are involved somehow!
2. Can a space probe run on open source software? It should be running Linux! No, FreeBSD! No, Linux! No, OpenBSD! No, Linux! No, the BeOS!
3. Does the Martian surface contain any naked petrified hot grits?
4. Space probes? Hey, how about a Beowulf cluster of those things!
X10 is a protocol for controlling appliances via power lines. For example, if your computer was fitted with an X10 interface, you could use it to control any X10-aware devices on the same electrical system; including lightbulbs, toasters, TV's, air conditioners, whatever. Of course, you have to go to the grief of making those devices X10 aware; and usually also installing an isolator so that you can't control your neighbours lights, and he can't control yours.
But it could easily have been slave conditions. My H1B visa allowed me to work for this one specific company in the U.S., and no other. If they fired me, I had to leave the country within ten days. Ten days! Sell my car, arrange to move all my possessions internationally, deal with ending my rental lease, blah blah f***ing blah, all within ten days!
And eventually, this is what did happen. The company, as high-tech startups are wont to do, went belly-up, leaving me stranded. I applied to get a tourist visa. This is trivially easy to do from outside the country; but impossible from inside, due to the beauracracy. Basically, this is one of the many parts of the INS which has ground to a complete halt. While this was going on, I applied for a job with the local library system. It was a job they'd advertised a bunch of times before, but not been able to fill, as they hadn't been able to find anyone with the right skillset. I, however, matched perfectly. This is what H1B's are about -- no American qualified to fill the job, so give it to a foreigner. They were all set to pay me US$45K to US$61K, when we found out how hard the immgration stuff was going to be. It would take 3 to 4 months to process my visa, and this processing had to be completed before I could do any work at all for my new prospective employer. Understandably, they turned down the opportunity to hire me.
And that 3 to 4 months was at the good end of the year! A month later, visas ran out, meaning that there would be a six month wait for the new year (with a new visa allotment) to start, and then 3 to 4 months on top of that! Outside the Silicon Valley firms which are specifically geared up to import tech workers, what organisation has any idea of what its high-tech employee requirements are going to be three to ten months ahead?
There are more visas for this coming "year" (which starts in October), which means that there won't be the six month wait, but with the same number of beauracrats to process way more visas, guess what? That's right, that 3 to 4 month wait is going to blow way out.
So I had to leave the US, leaving behind all my friends. I should mention that my experience was in the area aerved by the California branch of the INS. It is the one completely snowed by the influx to Silicon Valley; and I understand that other regional INS's are more on top of things.
It is my opinion that these massive delays have been allowed to creep in by three things: first, of course, the rapid increase in influx of workers; second in the US immigration law's attitude that America is the greatest country in the world and that if we didn't have these laws everybody in the world, all six bilion of them, would be on our doorstep tomorrow; and third in the fact that the people the INS serves generally aren't voters.
That's the limiting factor. You're right; obviously, you couldn't fit Linux in 2K words of code. So you have to expand the available code space. This is of course impossible -- the PIC is only capable of executing instructions from its internal store.
But it's possible to hang external memory off a PIC, whether it be a Dallas iButton or a more traditional RAM or ROM chip; and if you tried, you could probably build an interpreter to fit in the PICs internal instruction space, to execute "instructions" stored in this off-chip memory.
So yes, you could build a machine that ran Linux using a PIC as the processor; but Linux wouldn't be actually on board the PIC itself. Still, there are some nice small SMT EEPROM chips around, and you could fit one of those plus an SMT PIC on a pretty damn small board...
Incidentally, some of the newer PICs (e.g. the PIC16C77, I think) are not OTP; they're flash. Sweet.
I totally agree; I think the only real solution is to involve a human on an ongoing basis.
Note that to do this, you'll need to buy blocking software that allows you to edit the block list. I believe NetNanny is one such.
In the simple version, it would boot off CD, and look for local fd drives and hd partitions to auto-mount. The all-singing all-dancing version would scan those partitions, and look for a Windows registry with clues as to what name and IP number it should be; and wander out on the LAN and look for SMB shares to mount. It could come with tools to roll a new CD, with new apps etc.
Maybe it should be called Ronix, for Read-Only Linux.
Does anybody know if anything like this already exists?
It's because OpenBSD is the security-conscious version of BSD. They focus on having really good, really thorough security first, everything else second. As such, encrypting swap is a good, clever idea.
Recall that a Cat5 cable is usually:
1 - whiteOrange - 1
2 - Orange - 2
3 - whiteBlue - 3
4 - Green - 4
5 - whiteGreen - 5
6 - Blue - 6
7 - whiteBrown - 7
8 - Brown - 8
Important points:
all pins connect "straight": 1 to 1, 2 to 2, etc.
1 & 2 are a pair; and 3 & 6 are a pair.
4 & 5 and 7 & 8 are pairs too. They're not used by Cat5, 100Mbit; but they are used by Cat5e, 1Gbit; so hook 'em up now to save trouble later.
Now, 1 & 2 are the "transmit" pair; think of one line as the signal, and the other as the ground return for that signal. Each line needs its own return in order to get the groovy benefits of the magic of twisted pair. Similarly, 3 & 6 are the "receive" pair. When you hook a PC to a hub (the usual state of affairs), the "receive" and "transmit" are the other way 'round, so the PC transmits to the hub's receive, and vice versa, so all is happiness.
But when you're hooking two PC's together direct, if you used a straight cable, you'd be hooking one PC's transmit to the other PC's transmit, and the same for receive. No workee. So we swap those pairs:
1 - whiteOrange - 3
2 - Orange - 6
3 - whiteBlue - 1
4 - Green - 4
5 - whiteGreen - 5
6 - Blue - 3
7 - whiteBrown - 7
8 - Brown - 8
So that's 1 -> 3, 2 -> 6, 3 -> 1, 6 -> 2.
4 & 5 and 7 & 8 are again not used, but hook 'em up anyway.
When you test your cable, you can buy a cheap cable checker that shows a little light for each line (try Weidmuller / Paladin Tools), or a multimeter, and this will tell you whether you have continuity on each line. However, to test the cable properly, you really need a much more expensive checker that tests if it's gonna work at 100Mbit. After all, at 100Mbit, each bit is only 3 metres long!
It's great as the premise for an SF novel (and it is a really good novel); but it's bogus in real life.
In the Windows space, if you want to see code you wrote running on thousands of machines, one of the quickest ways up is to write a virus. But in the Linux space, the quickest way up is to join an open-source project. Not only do you get to point to some tweaky little feature in KDE or whatever and say, "I wrote that!"; but you're also (generally) made to feel welcome and supported by your fellow project members.
While they understand that it's illegal, I don't think they regard it as immoral. I used to be this way myself: way back when I owned a C-64, I used to pirate software because (a) I had no money, (b) everybody else was doing it, and (c) since I couldn't afford to buy the games anyway, I wasn't really taking money from their pocket, was I?
The immorality of removing someone's opportunity to make profit off their hard work is harder to see, particularly when everyone else is doing it, and particularly when you know that the original artists receive only a tiny share of the money anyhow).
I think that if music publishers made music available in a more direct way, and made a more direct pitch to the people doing the copying to please respect the rights of the artists (as opposed to jumping up and down and shouting, "Pirates are evil!"), that they might well find a satisfactory resolution.
For example, if the page is bloated, demo it on a dial-up line, so it loads really slowly. Emphasise to them that this is the way 90% of their customers will experience it.
Have a user who is unfamiliar with the stupid-mouseover thing sitting in the driver's seat when you demo it. Tell him to walk in a particular direction which will highlight the stupidness of said "feature".
OK, so it isn't always easy to demo stuff in the way you want, but sometimes it's possible.
Assuming AOL/TW are telling the truth, of course.
Technology defines society. For example, it used to be perfectly acceptable to keep slaves, because farming and manufacturing were labour-intensive. It was only once it was economically feasible to do away with slavery that the bulk of public opinion shifted, because until then the people in power had an interest in it not shifting.
Then, too, society defines technology. The internet itself is a creation of the U.S. government's tendency to throw vast numbers of dollars at military projects. Same deal for modems.
So one of the questions Katz is trying to ask here is: What technology should we create, in order to shape society?
Jack William Bell (who wins the "Most Karma gained in a Day" prize) makes the point that this is far from the whole story, and that law and propaganda are extremely important. For example, you and I will always be able to get hold of a copy of DeCSS, regardless of what various laws may say, because we have a clue; but if it's illegal, then it can't be included in your facourite Linux distro, and thus will never be available to the bulk of the population. Freedom for a small elite is a pretty sucky kind of freedom.
But inroads can be made through a technological front. As Dr Caleb has mentioned, the Internet exists in places other than the US.
Here is one such technological front: a data haven. A bunch of gutsy servers located technically, physically or practically outside the jurisdictions that want to control the Internet. One such model is described in Neal Stephenson's excellent Cryptonomicon; another is being created by people building a free networking system that caches data anonymously, thus meaning that there's no-one to serve with an injunction (can somebody remember their URL?). Want DeCSS? Go to a data haven. (Want kiddie porn? Go to a data haven. Clearly, complete freedom has a bad side as well as a good side, but to an extent ya gotta accept both sides of the coin, or neither.)
And hey, is anyone other than me having trouble believing that Katz said "Yo"?
Imagine you have a piece of paper, with a coded message on it. You can easily copy the piece of paper (and the message), using a photocopier.
The point is: copying has nothing whatsoever to do with decryption.
Further, imagine that these pieces of paper with coded messages were your DVD collection. Your DVD player has to be able to decode the message to play the movie. If you want to be able to build your own DVD player ('cos the commercial ones suck), your player has to be able to decrypt the code, just like existing players.
Again, the point is: the ability to copy and the ability to decode are completely different. Decoding is necessary for playing; it's not necessary for copying.
We've been hearing recently about quantum cryptography, in which data is sent in an uncrackable format; and also about quantum computers, whose massively parallel computation ability might make current crypto trivially easy to break.
I know that people have had some success in the laboratory with both of these things; do you see them succeeding in the real world? How soon?
I'm fascinated that a number of hackers, with little in common other than a desire and an ability to fix a piece of the problem, can convene in an ad-hoc forum and fix the problem; probably before the problem's "owners" are even aware of its existence.
Is this the first time slashdot has been used in this way, as a real-time tool to coordinate a debugging effort?
Some of my friends at the University Computer Club of the University of Western Australia did a very similar thing. They had an amazingly antique terminal controller, which they had aquired in the same way they got most of their hardware: it was totally obselete, and therefore free.
But it didn't come with the keyboards; so they made keyboards themselves. Take three computer punch-cards, cut holes in the middle one, in exactly the fashion outlined on Jim's site, place aluminium foil "row" strips between the first and second card, and "column" strips between the second and third. Draw "keys" on the first card. Voila! Instant keyboard.
As the keyboard got older, it would get harder to use, and you'd have to bash the keys with a pen, and eventually throw it away and build a new one.
The fundamental problem is that US currency is so easy to copy. I have easily enough stuff lying around my office to produce reasonably realistic copies.
Software aimed at specifically recognising currency and stamps is foolish: it will only recognise certain kinds of currency and stamps; it won't reconise foreign stamps; and it won't recognise other paper instruments which we would rather not see forged (certificates, etc.)
Software aimed at making forgeries trackable is more thoughtful; but it has obvious privacy implications, and is potentially technically defeatable (as many readers have mentioned).
The fundamental solution is to make currency harder to forge. Australian currency notes, for example, are printed on a thin papery plastic instead of on paper; they have a piece of artwork partly printed on each side, so it is obvious if the artwork on the two sides is misaligned; and they have a transparent section, so it is obvious if it is printed on the wrong "paper". In a similar vein, the new US $20 note has a "color change" section that looks different when viewed from different angles.
Trying to fix the problem by limiting the technology in a thousand different scanners, printers and copiers is a bad approach: it's analogous to trying to cover for your lousy encryption by crippling everybody else's computer. The Right Thing is improve the technology in the money itself.