You do not want to carry around hardware and
your files with you. You want to maybe make a local copy of an ideal virtual machine, which has all your apps and your data and your computing environment. See my paper on
The Linux Personal Virtual Server
Do you think you could build a mechanical ethernet interface?
It might need to use some paper tape to buffer the incoming packets before routing them. But there is hardware that does that: Reperferator transmitter
My usual 'gateway' machine is not located in my house... here in the future, we don't have a central server in the house, each device has its own embedded processor. I keep an image of my personal Linux virtual machine on my "server" offsite (see "user-mode-linux project"), and run local copies of my world on whatever machine I happen to be using at the moment. A thinkpad in this case.
Actually, you can overclock a teletype quite easily. Just change the gearing from the motor. I 'clocked' this one up to 75 baud from the stock 45.5 baud. But that is a factory approved setting. I don't know how fast you can gear it before it explodes though.
In the teletype there are no electronics at all, however the power supply bridge for the high voltage loop does have four solid-state diodes in it. Must have been the equivalent of a 3.5 GHZ Pentium back then, in terms of high tech leading edge technology!
The real point of the article was to point out that the Mini-ITX board, because of its PC compatibility and the price breaks due to consumer use, is now the embedded board of choice for standalone appliance types of projects. I think this is significant because it marks a crossover of the consumer hardware into to the embedded space, and also because it allows Linux to run in a huge new set of non-desktop areas.
The example with the teletype was really just a way of contrasting an app that previously would have required an industrial controller type of board, and now can use a consumer board.
You can also do a Google search, I have a script which tells Google that I am an i-Mode terminal (a Japanese mobile phone), and it formats the results and pages in small easily printable HTML-1.0 chunks.
The weather info comes from MIT's weather server which still formats the weather service info as teletype output.
Yup, works fine. I even had it hooked up as an alarm clock, I had a cron job to print the news and weather at 7:30 AM every morning. In the original military cabinet, it was loud enough to wake people up upstairs, but then I put it into a quieter cabinet, and it was too quiet to wake us up anymore.
But we use it every day, our Yahoo calendar sends events to it by email (no I won't tell you the address), and we see a printout every morning of the next day's events, easy to tear off and take with you on your way out.
The original inspiration for this was from Tom Jennings' "World Power Systems" site. Check it out. I used his ASCII-baudot conversion routines.
You see, this machine's print mechanism was a military model, and it has fully loaded "stunt box", which is basically a set of mechanical regexp state machines sitting on the printer control rods, it has about six of these, that can recognize sequences of three or four characters, and closes a relay, so you can control things like motors, other teletypes, coffee pots, etc.
It's almost enough to handle an HTTP request, if I had a paper tape loop and reader (excuse me, a "transmission distributor").
Now SCO is going after the goverment. When they try to go after the Army, that will be just like when Joe McCarthy tried to claim there were commies in the Army. They called his bluff, and that destroyed him.
I hope the military notices that they are being extorted by these liars from SCO, and lays some serious legal whoop-ass on them.
Maybe SCO, being the obvious scum sucking thieves that they are, actually took a piece of GNU code, copied it, changed a few lines, and are now trying to claim they wrote the code first.
There is no reason at all to think they are incapable of comitting this kind of felony.
I suggest that we start this as a rumour; there is nothing wrong with simply posing it as a question, just like Johnny Cochran did in the OJ trial ("have you read the entire manual for this DNA sequencer machine?"). Merely asking the question puts doubt in peoples minds, exactly as SCO has done.
they should look at the fibers at the edges
on
Picking Up the Pieces
·
· Score: 2, Interesting
Instead of looking at what's printed on the shreds, they should just scan the edges of each shred with a microscope. The orientation of the fibers at the edge would form a signature which could be matched to other shreds like a fingerprint. It would require higher res scanning, but I bet it would give almost perfect results.
We recently ordered a new consumer desktop PC from Dell, and it turns out is is substantially quieter than most PC's I have used. They seem to be paying attention to noise factors, or else we just happened to get a machine with particularly quiet power supply fans...
in 1997 we had a company that had a multiplayer game with chat, and we could set it to run messages through babelfish on the server (or whatever AltaVista's translation service was).
We even did translation twice for some games, The idea was to simulate international business negotiation, so to make the communication harder, we'd have messages translated from english to german to english, to simulate a scenario where a merger between an english and german company had taken place.
For people who want to play with error correcting codes, you can get the source code for a Reed-Solomon decoder/encoder I wrote here:
https://sourceforge.net/projects/rscode/
The people who build Linux don't want SCO's code
on
What if SCO is Right?
·
· Score: 2, Insightful
The thing that makes SCO so evil in this is that the people who make Linux distirbutions don't want any proprietary code, and they take pains to avoid it. If there is some proprietary code in there by mistake, the just and correct response is simply to remove it, and if it is critical, replace it with "clean" code written from scratch, like 99% of the system already is.
It's not like some proprietary code vendor stole some of SCO's code. The GPL crowd makes a lot of effort to keep their code free. It's like the difference between hitting someone with a car on purpose or by accident. If you accidently do harm, and you have taken extremely careful precautions to avoid harming others, you should not be charged with the same kind of crime and be subject to the same penalties as someone who harms another on purpose.
SCO needs to be punished somehow in this whole affair. I imagine no one is buying their aging properietary flavor of unix anymore, so perhaps the best punishment is to drive them all the way out of business. That means any company which uses SCO legacy code should switch to something else (Linux, BSD) *today*!
I have a small web browser (for CHTML) that runs in Java MIDP, and has specifically been ported to the aJile systems chipset (by me). It's called Picrobrowser, and I have versions both on iAppli (iMode) phones, and on MIDP.
If anyone would bother to actually read Turing's paper where he describes the 'test', you would see that he was not proposing the test literally, but as a reductio-ad-absurdum argument.
The issue was that many people at that time (and many today) seem to have a religious belief that thinking cannot be implemented in any way except with a human biological brain. Turing could clearly see that the human brain was a computational engine, and he of course defined the concept of a universal computer. Thus, it was obvious to him that you could build an artificial intelligence.
His "test" was really a way of gently pointing out the absurdity of the arguments of people like Searle (who came much later), who would blindly deny that a machine could ever think.
Turing's point was, to paraphrase "look, if I give you a machine which is indistinguishable in every respect from a human, which you can talk to in depth on any subject of the arts or sciences, and you *still* don't call that intelligence, then you are just so wedged that there is no point in talking about this anymore".
He would be saddened I think, and slightly disgusted, to see people twisting the whole purpose of his little thought experiment to argue for the kind of ignorance and transparently idiotic rhetoric of the kind that Searle and other "critics" of artificial intelligence try to make.
"All right, let me see if I understand the logic of this correctly. We are going to ignore the
United Nations in order to make clear to Saddam We're going to wage war to preserve the UN's ability to avert war. The paramount principle is that the UN's word must be taken seriously, and if we have to subvert its word to guarantee that it is, then by gum, we will. Peace is too important not to take up arms to defend. Am I getting this right?
No, this isn't about enforcing the power of the United Nations. Who gave you that idea?
Re:REPEAT AFTER ME: XML IS NOT A FILE FORMAT
on
Office 2003 and XML
·
· Score: 1
I think everyone knows what they meant; an XML file format for document interchange. Don't be so pedantic.
The argument always goes "it costs so much money to research and develop a new drug, that the only way to pay for it is to give a monopoly to the company that invents it".
However, there is another option, which is to have the government fund the development of drugs and have them belong to the public.
But, you say, then it's communist. We need the incentives of capitalism to spur people to do the research.
The cost of granting a monopoly to a company for an important drug is phenomenal, if they make 10 billion dollars profit on it, that is money that is taken literally from the pockets of everyone in our society.
So, how about if we have committees of scientists who direct research for the government, and if a drug that they champion is successful, we write them a personal check for 10 million dollars. That is one percent of the cost of granting the monopoly to a company, and the incentive for the indivudual scientist is now huge compared to what they could make at a salaried position.
In other words, apply the reward directly to the scientists, not to the parasite business people who run the drug companies.
Loebner didn't 'outfox' Minsky. Minsky asked for a prize to get Loebner to revoke his prize, not for someone to win it. But such a simple distinction is lost on an epsilon-minus such as yourself.
The author of the article appears never to have read the article by Turing where he described the so-called 'test'. It is clear that Turing was a deep and subtle thinker way ahead of his time. If you read what he is saying in context, he is arguing that first and foremost, thought can be automated in the sense of a universal computer which can compute anything that a brain can. To his critics who said that this was somehow impossible, he created a reducto-ad-absurdum argument; he said look if you are talking to this machine and it is composing sonnets which are like Shakespeare, and you *still* can't say it's intelligent, then you are an idiot. He was not proposing that this was an objective test or a desirable thing to do, he was poking fun at idiots like the author of the Salon article.
Slashdot Mirror
You do not want to carry around hardware and your files with you. You want to maybe make a local copy of an ideal virtual machine, which has all your apps and your data and your computing environment. See my paper on The Linux Personal Virtual Server
Do you think you could build a mechanical ethernet interface?
It might need to use some paper tape to buffer the incoming packets before routing them. But there is hardware that does that:
Reperferator transmitter
My usual 'gateway' machine is not located in my house... here in the future, we don't have a central server in the house, each device has its own embedded processor. I keep an image of my personal Linux virtual machine on my "server" offsite (see "user-mode-linux project"), and run local copies of my world on whatever machine I happen to be using at the moment. A thinkpad in this case.
Actually, you can overclock a teletype quite easily. Just change the gearing from the motor. I 'clocked' this one up to 75 baud from the stock 45.5 baud. But that is a factory approved setting. I don't know how fast you can gear it before it explodes though.
In the teletype there are no electronics at all, however the power supply bridge for the high voltage loop does have four solid-state diodes in it. Must have been the equivalent of a 3.5 GHZ Pentium back then, in terms of high tech leading edge technology!
The real point of the article was to point out that the Mini-ITX board, because of its PC compatibility and the price breaks due to consumer use, is now the embedded board of choice for standalone appliance types of projects. I think this is significant because it marks a crossover of the consumer hardware into to the embedded space, and also because it allows Linux to run in a huge new set of non-desktop areas.
The example with the teletype was really just a way of contrasting an app that previously would have required an industrial controller type of board, and now can use a consumer board.
The weather info comes from MIT's weather server which still formats the weather service info as teletype output.
But we use it every day, our Yahoo calendar sends events to it by email (no I won't tell you the address), and we see a printout every morning of the next day's events, easy to tear off and take with you on your way out.
The original inspiration for this was from Tom Jennings' "World Power Systems" site. Check it out. I used his ASCII-baudot conversion routines.
My God, man, that's a fantastic idea!
You see, this machine's print mechanism was a military model, and it has fully loaded "stunt box", which is basically a set of
mechanical regexp state machines sitting on the printer control rods, it has about six of these, that can recognize sequences of
three or four characters, and closes a relay, so you can control things like motors, other teletypes, coffee pots, etc.
It's almost enough to handle an HTTP request, if I had a paper tape loop and reader (excuse me, a "transmission distributor").
Now SCO is going after the goverment. When they try to go after the Army, that will be just like
when Joe McCarthy tried to claim there were commies in the Army. They called his bluff, and that destroyed him.
I hope the military notices that they are being extorted by these liars from SCO, and lays some serious legal whoop-ass on them.
Maybe SCO, being the obvious scum sucking thieves that they are, actually took a piece of GNU code, copied it, changed a few lines, and are now trying to claim they wrote the code first.
There is no reason at all to think they are incapable of comitting this kind of felony.
I suggest that we start this as a rumour; there is nothing wrong with simply posing it as a question, just like Johnny Cochran did in the OJ trial ("have you read the entire manual for this DNA sequencer machine?"). Merely asking the question puts doubt in peoples minds, exactly as SCO has done.
Instead of looking at what's printed on the shreds, they should just scan the edges of each shred with a microscope. The orientation of the fibers at the edge would form a signature which could be matched to other shreds like a fingerprint. It would require higher res scanning, but I bet it would give almost perfect results.
We recently ordered a new consumer desktop PC from Dell, and it turns out is is substantially quieter than most PC's I have used. They seem to be paying attention to noise factors, or else we just happened to get a machine with particularly quiet power supply fans...
in 1997 we had a company that had a multiplayer
game with chat, and we could set it to run messages
through babelfish on the server (or whatever AltaVista's translation service was).
We even did translation twice for some games, The idea was to simulate international business negotiation, so to make the communication harder, we'd have messages translated from english to german to english, to simulate a scenario where a merger between an english and german company had taken place.
For people who want to play with error correcting codes, you can get the source code for a Reed-Solomon decoder/encoder I wrote here:
https://sourceforge.net/projects/rscode/
The thing that makes SCO so evil in this is that the people who make Linux distirbutions don't want any proprietary code, and they take pains to avoid it. If there is some proprietary code in there by mistake, the just and correct response is simply to remove it, and if it is critical, replace it with "clean" code written from scratch, like 99% of the system already is.
It's not like some proprietary code vendor stole some of SCO's code. The GPL crowd makes a lot of effort to keep their code free. It's like the difference between hitting someone with a car on purpose or by accident. If you accidently do harm, and you have taken extremely careful precautions to avoid harming others, you should not be charged with the same kind of crime and be subject to the same penalties as someone who harms another on purpose.
SCO needs to be punished somehow in this whole affair. I imagine no one is buying their aging properietary flavor of unix anymore, so perhaps the best punishment is to drive them all the way out of business. That means any company which uses SCO legacy code should switch to something else (Linux, BSD) *today*!
See http://sourceforge.net/projects/bearlib/ see the PicoMIDP project.
If anyone would bother to actually read Turing's paper where he describes the 'test', you would see that he was not proposing the test literally, but as a reductio-ad-absurdum argument.
The issue was that many people at that time (and many today) seem to have a religious belief that thinking cannot be implemented in any way except with a human biological brain. Turing could clearly see that the human brain was a computational engine, and he of course defined the concept of a universal computer. Thus, it was obvious to him that you could build an artificial intelligence.
His "test" was really a way of gently pointing out the absurdity of the arguments of people like Searle (who came much later), who would blindly deny that a machine could ever think.
Turing's point was, to paraphrase "look, if I give you a machine which is indistinguishable in every respect from a human, which you can talk to in depth on any subject of the arts or sciences, and you *still* don't call that intelligence, then you are just so wedged that there is no point in talking about this anymore".
He would be saddened I think, and slightly disgusted, to see people twisting the whole purpose of his little thought experiment to argue for the kind of ignorance and transparently idiotic rhetoric of the kind that Searle and other "critics" of artificial intelligence try to make.
No, this isn't about enforcing the power of the United Nations. Who gave you that idea?
I think everyone knows what they meant; an XML file format for document interchange. Don't be so pedantic.
The argument always goes "it costs so much money to research and develop a new drug, that the only way to pay for it is to give a monopoly to the company that invents it".
However, there is another option, which is to have the government fund the development of drugs and have them belong to the public.
But, you say, then it's communist. We need the incentives of capitalism to spur people to do the research.
The cost of granting a monopoly to a company for an important drug is phenomenal, if they make 10 billion dollars profit on it, that is money that is taken literally from the pockets of everyone in our society.
So, how about if we have committees of scientists who direct research for the government, and if a drug that they champion is successful, we write them a personal check for 10 million dollars. That is one percent of the cost of granting the monopoly to a company, and the incentive for the indivudual scientist is now huge compared to what they could make at a salaried position.
In other words, apply the reward directly to the scientists, not to the parasite business people who run the drug companies.
http://www.laszlosystems.com
It is a compiler that produces Flash apps, much easier to program than Macromedia's ActionScript
tools.
Loebner didn't 'outfox' Minsky. Minsky asked for a prize to get Loebner to revoke his prize, not for someone to win it. But such a simple distinction is lost on an epsilon-minus such as yourself.
for cancer for the last thirty years. Frauds!
The author of the article appears never to have read the article by Turing where he described the so-called 'test'. It is clear that Turing was a deep and subtle thinker way ahead of his time. If you read what he is saying in context, he is arguing that first and foremost, thought can be automated in the sense of a universal computer which can compute anything that a brain can. To his critics who said that this was somehow impossible, he created a reducto-ad-absurdum argument; he said look if you are talking to this machine and it is composing sonnets which are like Shakespeare, and you *still* can't say it's intelligent, then you are an idiot. He was not proposing that this was an objective test or a desirable thing to do, he was poking fun at idiots like the author of the Salon article.