Why is the article icon AMD's? I don't see any relation to the scrappy little semiconductor company in any of this. If anything, I would expect that we'd have the Christmas Tree icon here, or the Technology icon.
The first GSM mobile phone I ever owned (back in 1999), a Nokia 3210, had a ring tone composer which I could use to send ring tones I composed to my friends who had compatible phones. Newer 3xxx (e.g. 3310/3350, etc.) models even have the ability to resend tones that have been received. Heck, if this isn't (an admittedly primitive) P2P network built on top of GSM, I don't know what one is. With SMS chat services, getting the tones you want is not too difficult.
But then again, it seems that the United States is somewhat backwards when it comes to cellular telephony for some reason. We've been doing this in the Philippines for at least five years almost.
Odd thing for a third world country like us to have market penetration rates for cellular phones approaching that of the wealthiest European nations. Heck, I see street vendors here who have GSM mobiles!
This isn't the first time Apple (or at any rate its returned founder Steve Jobs) has had a run-in with the GPL. NeXT used GCC to speed up their development of an Objective-C system, hoping that there was a loophole in the GPL that would allow them to do creative linking of binary object files that they would keep proprietary, to provide the Objective-C support. The FSF, however, was wise to this and told them that no, they can't do that, and after some legal wrangling, NeXT realized they were wrong, and opened up the source for the Objective-C compiler. This is also the main reason why gcc is probably the only compiler in the world for Objective-C (not counting the to-C translators out there), and why it's the default compiler system on MacOS X (thanks to Cocoa's lineage from NeXTSTEP).
This has been a Good Thing(tm) for the rest of the world, but to someone like Steve Jobs who isn't convinced of the merits of open development and Free Software, it would have felt like the Free Software Foundation got one up on him.
Correction is in order here though. The GPL erroneously described to give restrictions on developers. Nonsense. The GPL places restrictions on DISTRIBUTORS of software. You can make whatever modifications you please to GPLed software to your heart's content, and no one will call on you as long as it stays internal. The GPL only comes into play when you redistribute these modifications to third parties. You have to provide these third parties with your (modified) sources, and you cannot thereafter legally restrict them from in turn modifying your sources and/or redistributing them to anyone they see fit.
Re:Encryption and compression make a lot of sense.
on
PKWare Zips to Growth
·
· Score: 5, Insightful
Hopefully, if this is what they want to do, they will do better than the embarrasingly insecure "encryption" that the old DOS PKZip included (a cryptographically-weak LFSR-based stream cipher). With good support for cryptographic standards, they could have something here.
By the way, you always do encryption AFTER data compression. Doing it before data compression ensures that your compression ratio is close to 0%.
Just a slight nitpick, but AFAIK, Kerberos never used any public key cryptography at all, Diffie-Hellman or otherwise. They use the Needham-Schroeder key exchange protocol which only requires symmetric key cryptography.
I wonder if they have any of the anomalous 16th century maps that happen to accurately depict Antarctica as though it were free of ice. That's of course impossible, but there are maps, the most famous being the Admiral Piri Reis map, that accurately display the subglacial topography. There are also maps by Mercator and Buache that also display antarctic subglacial features.
Of course, the subglacial topography of Antarctica was unknown until sonar surveys of the 1950's, and the whole continent itself was unknown until the early 19th century.
Gee, that's an explosive way to counteract the effects of acid rain! You'd get sodium sulfate in the water, but I imagine it's not nearly as bad as sulfuric acid.
Of course, as with all "breaks" of cryptographic algorithms out there, the Courtois-Pieprzyk XL/XSL attack on AES was nothing but an academic break. From what I heard of it, to break a 128-bit AES key you still need to do approximately 2^100 encryptions, 1.26e30 encryptions, which is impossible even for the NSA. For Serpent (which is still widely considered to be the AES candidate with the highest security margin), the 256-bit key would still require something like 2^200 encryptions, still impossible unless you could get every sub-atomic particle in the universe to do a billion encryptions every second! I think the same is true if you had AES with the full key strength.
No, even if the breaks were true, I'd still be confident in the security of my AES-encrypted files. I'd start thinking of other alternatives, sure, but I won't go back to using triple-DES.
Of course Hawass and his ilk are disdainful of this theory, but in the Sphinx, there is reportedly a hidden "hall of records" that is supposed to be under one of the shoulders of the Sphinx. According to Edgar Cayce, it was supposed to crumble away around the year 2000 and humankind's lost knowledge would be revealed. Naturally, Hawass and his ilk conducted extensive repairs of the Sphinx a year or two before Cayce's prophecy was supposed to come true, although there have been ground penetrating radar surveys that showed that there is indeed something there. Oddly enough, Mark Lehner was once upon a time associated with Cayce's estate and was a member of one of these teams led by SRI International that discovered this tantalizing evidence. What brought about his volte face so he's now 100% with the orthodox Egyptological establishment is unknown.
But they aren't going to the Sphinx, right? They're shooting for that strange sealed door at the end of one of the shafts inside the Queen's Chamber. After Rudolf Gantenbrink discovered this mysterious door in 1993, which also had a wood fragment in front of it which Gantenbrink's robot couldn't pick up (and settle once and for all the question of when the Pyramid had been built), he was prevented from sending yet another improved robot down the shaft. Political pissing contest at the Antiquities Bureau sounded like. If they do manage to get the wood fragment, hopefully there'll be no cover up on the radiocarbon dating results. There has already been one (on the mortar used in some parts of the Pyramid, conducted in 1986) that gives a date for the building of the Pyramid about a thousand or so years before the 2500 BC date traditionally accepted by Egyptologists. Gee, so the Pyramid was supposedly built ca. 3800 to 2800 BC. In the Predynastic Period! Of course these results were universally ignored by Egyptologists as it would greatly upset the neat chronology they have for the building of the Giza monuments.
This is what I get from reading too much Graham Hancock!
The software is built in a similar way - lots of internal checks, tell-me-thrice memory, soft-failure-bit-flip-correcting daemons etc. In this case, lives aren't at stake, but the people doing the programming are used to situations where they are.
Not only that, a single space launch of even a fairly small satellite still costs over a billion dollars. If there's a software glitch, it could render the satellite totally inoperable, and I doubt that these engineers want to tell their source of funding that a glitch they're responsible for just wasted the whole launch...
Which is also why Microsoft doesn't do aerospace embedded systems.:) Whoops, Satellite Redmond I just had a BSOD...
Well, in an ironic twist of fate the late Ozamu Tezuka was himself ripped off by Disney after his death, despite his wife's protestations to the contrary (bought out by the Mouse perhaps?).
Or has no one else noticed how blatantly similar Tezuka's "Kimba the White Lion" was to Disney's "The Lion King"? Or that "Atlantis the Lost Empire" is so disturbingly similar to GAINAX's "Nadia: The Secret of Blue Water" to be the most obvious candidate for true Disney plagiarism?
What Schneier actually advocates in the article is the use of at least two of these three layers for doing user authentication: something you know (e.g. a password), something you have (e.g. a smart card or other secure token), and something you are (biometrics falls into this rubric). Depending on only one is necessarily weak, but even two of the three taken together would be strong indeed. For instance, if you have a website that uses not only username/password pairs for authentication, but lives on SSL *and* requests client-side certificates from any browser that wishes to visit the protected page uses both something you know (your username and password) and something you have (the computer where the browser with the client-side certificate is installed, or better yet if the cert lives on a smart card). THAT would make Schneier's Parable of the Dirty Website fail utterly without extra work: without the client-side cert, the web page wouldn't even serve the username/password page to you. Fine, the password is compromised because the employee used the same password to surf for porn, but since access to the certificate is limited to the computer where it's installed, or the smart card possessed by the employee no dice unless you can also steal the smart card and/or computer. Even better would be to provide biometric authentication for the secured computer, so you'd then have to steal the fingerprint or retinal scan or whatnot as well to break the system.
It can be done of course, but it would require contortions worthy of Sneakers.
The whole article actually feels like a distillation of the last six months of the Crypto-Gram newsletter.
I believe the Internet core routing protocols use Dijkstra's shortest path algorithm, whereas RTS games probably use the A* algorithm to find approximate shortest paths. So everyone who accesses slashdot remotely uses his algorithm...:) IIRC, Dijkstra also developed semaphores and mutexes, according to our old friend Andy Tannenbaum, which are an absolute requirement for any multitasking, multithreaded OS. Gosh, the man was a legend...
How can you actually "see" stuff at quantum scale?
on
Java Powers of Ten
·
· Score: 5, Interesting
The short answer is, I guess, you can't. Quantum objects like molecules, atoms, and sub-atomic particles will always be "invisible", as they are all much, much smaller than a wavelength of visible light, which is what we really define vision as. We can really only infer their existence from their indirect effects, which is the only way we know any of them are real. Besides, to actually "see" anything amounts to measuring the position and velocity of an object to as high an accuracy as the size of the object, so the Heisenberg uncertainty principle makes it impossible to see anything so small...
An attempt to actually zoom into a proton to see it using high-energy gamma ray photons would require a photon wavelength of less than 1 fm, or about 10^23 Hz. This gives a photon energy of roughly 2.5 GeV, which is comparable to the energies generated at the Fermilab or CERN particle accelerators. I guess this is probably enough energy to turn the proton into something else entirely even before you could see it. A similar attempt to view an atom would require a photon wavelength of 1 angstrom, a wavelength of about 10^18 Hz, and a photon energy of about 12 keV, quite enough to completely ionize the atom and strip away all of its electrons, leaving you with nothing to see. A similar calculation for the DNA strands at 10^-9 m gives an approximately 124 eV photon energy, which is also sufficient to ionize some of the molecules; you may be able to get a picture, but it will be a very hazy one (the best electron microscopy has been able to just barely make out the double helix structure of DNA).
You're right on almost all points, however, I think that you have gotten things slightly wrong with point #1:
1. Because the user has access to the source code, it's possible for him to make the modifications himself. In fact, the GPL encourages this. So chances are, he won't pay someone else to do it.
Most of the companies who need specific modifications are generally unable or unwilling to perform these modifications themselves. If you have one of those companies which are not in the business of actually writing and creating software (and of course there are far more of these than those that do), for them to actually make customizations of some GPLed software for themselves is counterproductive and wasteful. Yeah, if you think of it this way, they'd pay someone else to do the work I think.
Programming is not easy work, and understanding and making modifications to someone else's code is even harder. Have you ever tried to make major modifications to a large free software codebase? If so, then maybe you'll understand why most enterprises are reluctant to do this kind of work themselves, and prefer to outsource such work to other companies that specialize in the task.
Part of the problem has to do with the fact that Moore's law is still true, and hardware manufacturers everywhere are striving ever harder to extend its lifetime more than half a generation after Gordon Moore himself said it would end. This means that technology is getting more and more complex at an exponential rate, so much so that only those of us who are comfortable with such rapid evolution can stay on top of things, and usually, not even then. This is an explosion of complexity, and I think most people do not get a buzz out of comprehending complexity that changes all the time. In fact, such an explosion of complexity is terrifying to almost everyone, and is at the core of most technophobia.
Moore's Law should come to an end, and by this I mean that the pace of evolutionary progress of semiconductor and hence computer technology would slow down to perhaps a linear scale, or any scale that would give time for most people to think about what they have and what they can do with it. That's the problem with Moore's Law: it doesn't give anyone but those who choose to be part of the technological élite to even think about how it will affect their lives.
Moore's Law, software bloat, and the market
on
Linux on Older Hardware
·
· Score: 4, Insightful
Oh, how I wish Moore's Law would finally come to an end soon, or at least come to the point that it becomes impractical for the mass market to bear the cost of supporting its continued geometric growth. The factuality of Moore's Law is one of the biggest problems with the computer market: it's truth means that the market is not stable. This allows software makers to become sloppy with their design decisions because they wind up thinking, "Oh, it's slow now, but in 18 months the top of the line systems will double in power and then have enough computing power to run this kind of bloated crap I'm putting out without being as slow as a tired snail." It's as much true of the mainstream Linux distro makers as much as it is true of Microsoft and other proprietary software vendors.
Just for my workaday Linux distro, Red Hat 7.1. I for the life of me cannot understand why in heaven's name I need to install Kerberos to install the RPM package for CVS or LPRng. I don't have a Kerberized network and have no intention of setting such a creature up anytime in the near future, and likely it's true for most everyone. Or why I'm forced to install Japanese TTF fonts (xtt-fonts) just to get GhostScript up and running, or why printconf has to have a Kanji converter (nkf). I don't read Japanese, and I imagine the vast majority of the users of Red Hat's standard edition will never have any need to view, much less print, a Japanese-language document. The list of odd dependencies can go on and on ad nauseam, and there are many other signs of bloat. It's this kind of bloat that makes it impossible to run an up to date Linux distro on older hardware.
The other problem comes from hardware manufacturers, which is why unless Moore's Law comes to an end someday, this trend is going to keep going. And never mind us folks whose incomes cannot support a major hardware upgrade every 18 months. When a new technology appears, they stop making the old technology almost instantly. Can you still buy EDO SIMM's? Can you still buy a non-AGP video card? Well, unless you go to a surplus shop, probably not. Because of Moore's Law and its effect on the market, obsolete hardware has a way of becoming impractical or even impossible to maintain at some point, which is why everyone, even us in the third world who don't have a lot of disposable income and can't constantly support hardware upgrades, is eventually forced to upgrade.
While this project's aims are commendable, I wouldn't hold out too much hope for a universal adoption of its philosophy, not until Moore's Law comes to an end and the computer hardware market stabilizes as a result. Until then, I hope they remain true to the vision and not succumb to the temptations that have created the bloated monstrosities common nowadays.
Tell me about it. The authors (the so-called "Gang of Four") of the book Design Patterns: Elements of Reusable Object-Oriented Software drew on the work of the architect Christopher Alexander for their seminal work in object oriented design. You might even go so far as to say that the purest fusion of art and technology is really architecture in this more general sense.
Why is it everytime Cringely has a new article...
on
Cringely's Bank Shot
·
· Score: 2
...the PBS site goes down? It usually takes a week after he writes something new before I can see it. Is it because www.pbs.org is some dinky old server that takes a week to get up again after a slashdotting? Lately Slashdot's been linking to every article he writes, so of course, everyone's gonna try to look at it. This behavior has been consisten since I started reading Cringely's articles regularly last year.
Three VM's for the Open Sourcers under the sky
Seven VM's for the chipmakers with their foundries of stone
Nine for the mass market doomed to die
One for the Dark Lord on his dark throne
In the Land of Redmond where the Shadows Lie
One Runtime to Rule them All, One Runtime to find them
One Runtime to bring them and with.NET bind them
In the Land of Redmond where the Shadows Lie
Gee, so I guess that makes Miguel de Icaza Celebrimor, building his own runtime based on secrets given to him by the Dark Lord of Redmond, disguised as Annatar, Lord of Gifts. Maybe RMS is Elrond, watchful and distrustful of this mysterious being bearing secrets...
Re:Has he talked about Rod Brooks?
on
Arguing A.I.
·
· Score: 2
Well, coming from the point of view of traditional AI research it is truly radical. Call that the straitjacketed minds of crusty philosophers stuck in the ivory towers of academe, caught up in the biases of Western thought that seeks to divide, compartmentalize, and analyze the system of the world to understand it!
Brooks himself got these ideas from biology, a study so very far removed from the fields of computer science and electrical engineering that form the core of traditional AI research. It was only by stepping outside the bounds of traditionalist Western ideas about the compartmentalization of learning and knowledge that he brought these ideas forth.
I wonder what other ideas might come from a more integrated view of science, as opposed to the divisive approach Western science has taken.
Re:Has he talked about Rod Brooks?
on
Arguing A.I.
·
· Score: 2
Right. But then, a 'creature' with those kinds of 'sense organs' would be completely different and utterly alien to us. Because its experience of the world is utterly different, its emergent behavior would also be utterly different. In order to create a human-like creature, with human-like intelligence, the holy grail of AI, it would then necessarily have to have the sensory capabilities that a human would have. Either that, or you model the entire environment a human normally interacts with and allow your "artificially intelligent" being to interact with that simulated world, which is what traditional AI is trying to do. Of course, it's almost completely impossible to do that in its fullest generality... Very old arguments put forth by Hubert Dreyfus and Joseph Weizenbaum in the mid-1960's.
Has he talked about Rod Brooks?
on
Arguing A.I.
·
· Score: 3, Informative
I wonder if he talks about Professor Rodney A. Brooks at MIT and his ideas about artificial intelligence, situatedness, and embodiment.
For Rod Brooks, "intelligence" cannot really be programmed into a system; it is rather an emergent property of systems as they interact with their environment. In The Matrix Morpheus says that the body cannot exist without the mind, but Brooks would rather say that the mind cannot exist without the body, because the body is the only way that the mind can have any experience of its environment. It's a radical idea. It answers the problems behind knowledge representation that have been argued by Hubert Dreyfus in 1965, where he stated that any representation of knowledge is incomplete without its connection to all other pieces of knowledge. The paradigm Brooks is presenting in his ideas about embodied intelligence is that explicit representation of knowledge is superfluous: let the world itself be its own best model, and let the artificially intelligent being formulate its own judgments about what the world is and what it means from its own experience of that world. Intelligence emerges from its interaction and experience of the world. If Brooks is correct, then true AI is absolutely inseperable from robotics.
The seminal paper where Brooks discusses this philosophy is "Intelligence Without Reason" and is available at his website which is linked above.
Any book on AI that does not discuss this other branch of AI philosophy is in my view hopelessly incomplete.
Slashdot Mirror
Most of these companies actually lobbied for the DMCA when it came out! Oh the hypocrisy of it all...
Why is the article icon AMD's? I don't see any relation to the scrappy little semiconductor company in any of this. If anything, I would expect that we'd have the Christmas Tree icon here, or the Technology icon.
The first GSM mobile phone I ever owned (back in 1999), a Nokia 3210, had a ring tone composer which I could use to send ring tones I composed to my friends who had compatible phones. Newer 3xxx (e.g. 3310/3350, etc.) models even have the ability to resend tones that have been received. Heck, if this isn't (an admittedly primitive) P2P network built on top of GSM, I don't know what one is. With SMS chat services, getting the tones you want is not too difficult.
But then again, it seems that the United States is somewhat backwards when it comes to cellular telephony for some reason. We've been doing this in the Philippines for at least five years almost.
Odd thing for a third world country like us to have market penetration rates for cellular phones approaching that of the wealthiest European nations. Heck, I see street vendors here who have GSM mobiles!
This isn't the first time Apple (or at any rate its returned founder Steve Jobs) has had a run-in with the GPL. NeXT used GCC to speed up their development of an Objective-C system, hoping that there was a loophole in the GPL that would allow them to do creative linking of binary object files that they would keep proprietary, to provide the Objective-C support. The FSF, however, was wise to this and told them that no, they can't do that, and after some legal wrangling, NeXT realized they were wrong, and opened up the source for the Objective-C compiler. This is also the main reason why gcc is probably the only compiler in the world for Objective-C (not counting the to-C translators out there), and why it's the default compiler system on MacOS X (thanks to Cocoa's lineage from NeXTSTEP).
This has been a Good Thing(tm) for the rest of the world, but to someone like Steve Jobs who isn't convinced of the merits of open development and Free Software, it would have felt like the Free Software Foundation got one up on him.
Correction is in order here though. The GPL erroneously described to give restrictions on developers. Nonsense. The GPL places restrictions on DISTRIBUTORS of software. You can make whatever modifications you please to GPLed software to your heart's content, and no one will call on you as long as it stays internal. The GPL only comes into play when you redistribute these modifications to third parties. You have to provide these third parties with your (modified) sources, and you cannot thereafter legally restrict them from in turn modifying your sources and/or redistributing them to anyone they see fit.
Hopefully, if this is what they want to do, they will do better than the embarrasingly insecure "encryption" that the old DOS PKZip included (a cryptographically-weak LFSR-based stream cipher). With good support for cryptographic standards, they could have something here.
By the way, you always do encryption AFTER data compression. Doing it before data compression ensures that your compression ratio is close to 0%.
Golden parachute.
Just a slight nitpick, but AFAIK, Kerberos never used any public key cryptography at all, Diffie-Hellman or otherwise. They use the Needham-Schroeder key exchange protocol which only requires symmetric key cryptography.
I wonder if they have any of the anomalous 16th century maps that happen to accurately depict Antarctica as though it were free of ice. That's of course impossible, but there are maps, the most famous being the Admiral Piri Reis map, that accurately display the subglacial topography. There are also maps by Mercator and Buache that also display antarctic subglacial features.
Of course, the subglacial topography of Antarctica was unknown until sonar surveys of the 1950's, and the whole continent itself was unknown until the early 19th century.
Gee, that's an explosive way to counteract the effects of acid rain! You'd get sodium sulfate in the water, but I imagine it's not nearly as bad as sulfuric acid.
Of course, as with all "breaks" of cryptographic algorithms out there, the Courtois-Pieprzyk XL/XSL attack on AES was nothing but an academic break. From what I heard of it, to break a 128-bit AES key you still need to do approximately 2^100 encryptions, 1.26e30 encryptions, which is impossible even for the NSA. For Serpent (which is still widely considered to be the AES candidate with the highest security margin), the 256-bit key would still require something like 2^200 encryptions, still impossible unless you could get every sub-atomic particle in the universe to do a billion encryptions every second! I think the same is true if you had AES with the full key strength.
No, even if the breaks were true, I'd still be confident in the security of my AES-encrypted files. I'd start thinking of other alternatives, sure, but I won't go back to using triple-DES.
Of course Hawass and his ilk are disdainful of this theory, but in the Sphinx, there is reportedly a hidden "hall of records" that is supposed to be under one of the shoulders of the Sphinx. According to Edgar Cayce, it was supposed to crumble away around the year 2000 and humankind's lost knowledge would be revealed. Naturally, Hawass and his ilk conducted extensive repairs of the Sphinx a year or two before Cayce's prophecy was supposed to come true, although there have been ground penetrating radar surveys that showed that there is indeed something there. Oddly enough, Mark Lehner was once upon a time associated with Cayce's estate and was a member of one of these teams led by SRI International that discovered this tantalizing evidence. What brought about his volte face so he's now 100% with the orthodox Egyptological establishment is unknown.
But they aren't going to the Sphinx, right? They're shooting for that strange sealed door at the end of one of the shafts inside the Queen's Chamber. After Rudolf Gantenbrink discovered this mysterious door in 1993, which also had a wood fragment in front of it which Gantenbrink's robot couldn't pick up (and settle once and for all the question of when the Pyramid had been built), he was prevented from sending yet another improved robot down the shaft. Political pissing contest at the Antiquities Bureau sounded like. If they do manage to get the wood fragment, hopefully there'll be no cover up on the radiocarbon dating results. There has already been one (on the mortar used in some parts of the Pyramid, conducted in 1986) that gives a date for the building of the Pyramid about a thousand or so years before the 2500 BC date traditionally accepted by Egyptologists. Gee, so the Pyramid was supposedly built ca. 3800 to 2800 BC. In the Predynastic Period! Of course these results were universally ignored by Egyptologists as it would greatly upset the neat chronology they have for the building of the Giza monuments.
This is what I get from reading too much Graham Hancock!
Not only that, a single space launch of even a fairly small satellite still costs over a billion dollars. If there's a software glitch, it could render the satellite totally inoperable, and I doubt that these engineers want to tell their source of funding that a glitch they're responsible for just wasted the whole launch...
Which is also why Microsoft doesn't do aerospace embedded systems. :) Whoops, Satellite Redmond I just had a BSOD...
Well, in an ironic twist of fate the late Ozamu Tezuka was himself ripped off by Disney after his death, despite his wife's protestations to the contrary (bought out by the Mouse perhaps?).
Or has no one else noticed how blatantly similar Tezuka's "Kimba the White Lion" was to Disney's "The Lion King"? Or that "Atlantis the Lost Empire" is so disturbingly similar to GAINAX's "Nadia: The Secret of Blue Water" to be the most obvious candidate for true Disney plagiarism?
What Schneier actually advocates in the article is the use of at least two of these three layers for doing user authentication: something you know (e.g. a password), something you have (e.g. a smart card or other secure token), and something you are (biometrics falls into this rubric). Depending on only one is necessarily weak, but even two of the three taken together would be strong indeed. For instance, if you have a website that uses not only username/password pairs for authentication, but lives on SSL *and* requests client-side certificates from any browser that wishes to visit the protected page uses both something you know (your username and password) and something you have (the computer where the browser with the client-side certificate is installed, or better yet if the cert lives on a smart card). THAT would make Schneier's Parable of the Dirty Website fail utterly without extra work: without the client-side cert, the web page wouldn't even serve the username/password page to you. Fine, the password is compromised because the employee used the same password to surf for porn, but since access to the certificate is limited to the computer where it's installed, or the smart card possessed by the employee no dice unless you can also steal the smart card and/or computer. Even better would be to provide biometric authentication for the secured computer, so you'd then have to steal the fingerprint or retinal scan or whatnot as well to break the system.
It can be done of course, but it would require contortions worthy of Sneakers .
The whole article actually feels like a distillation of the last six months of the Crypto-Gram newsletter.
I believe the Internet core routing protocols use Dijkstra's shortest path algorithm, whereas RTS games probably use the A* algorithm to find approximate shortest paths. So everyone who accesses slashdot remotely uses his algorithm... :) IIRC, Dijkstra also developed semaphores and mutexes, according to our old friend Andy Tannenbaum, which are an absolute requirement for any multitasking, multithreaded OS. Gosh, the man was a legend...
The short answer is, I guess, you can't. Quantum objects like molecules, atoms, and sub-atomic particles will always be "invisible", as they are all much, much smaller than a wavelength of visible light, which is what we really define vision as. We can really only infer their existence from their indirect effects, which is the only way we know any of them are real. Besides, to actually "see" anything amounts to measuring the position and velocity of an object to as high an accuracy as the size of the object, so the Heisenberg uncertainty principle makes it impossible to see anything so small...
An attempt to actually zoom into a proton to see it using high-energy gamma ray photons would require a photon wavelength of less than 1 fm, or about 10^23 Hz. This gives a photon energy of roughly 2.5 GeV, which is comparable to the energies generated at the Fermilab or CERN particle accelerators. I guess this is probably enough energy to turn the proton into something else entirely even before you could see it. A similar attempt to view an atom would require a photon wavelength of 1 angstrom, a wavelength of about 10^18 Hz, and a photon energy of about 12 keV, quite enough to completely ionize the atom and strip away all of its electrons, leaving you with nothing to see. A similar calculation for the DNA strands at 10^-9 m gives an approximately 124 eV photon energy, which is also sufficient to ionize some of the molecules; you may be able to get a picture, but it will be a very hazy one (the best electron microscopy has been able to just barely make out the double helix structure of DNA).
You're right on almost all points, however, I think that you have gotten things slightly wrong with point #1:
Most of the companies who need specific modifications are generally unable or unwilling to perform these modifications themselves. If you have one of those companies which are not in the business of actually writing and creating software (and of course there are far more of these than those that do), for them to actually make customizations of some GPLed software for themselves is counterproductive and wasteful. Yeah, if you think of it this way, they'd pay someone else to do the work I think.
Programming is not easy work, and understanding and making modifications to someone else's code is even harder. Have you ever tried to make major modifications to a large free software codebase? If so, then maybe you'll understand why most enterprises are reluctant to do this kind of work themselves, and prefer to outsource such work to other companies that specialize in the task.
Part of the problem has to do with the fact that Moore's law is still true, and hardware manufacturers everywhere are striving ever harder to extend its lifetime more than half a generation after Gordon Moore himself said it would end. This means that technology is getting more and more complex at an exponential rate, so much so that only those of us who are comfortable with such rapid evolution can stay on top of things, and usually, not even then. This is an explosion of complexity, and I think most people do not get a buzz out of comprehending complexity that changes all the time. In fact, such an explosion of complexity is terrifying to almost everyone, and is at the core of most technophobia.
Moore's Law should come to an end, and by this I mean that the pace of evolutionary progress of semiconductor and hence computer technology would slow down to perhaps a linear scale, or any scale that would give time for most people to think about what they have and what they can do with it. That's the problem with Moore's Law: it doesn't give anyone but those who choose to be part of the technological élite to even think about how it will affect their lives.
Oh, how I wish Moore's Law would finally come to an end soon, or at least come to the point that it becomes impractical for the mass market to bear the cost of supporting its continued geometric growth. The factuality of Moore's Law is one of the biggest problems with the computer market: it's truth means that the market is not stable. This allows software makers to become sloppy with their design decisions because they wind up thinking, "Oh, it's slow now, but in 18 months the top of the line systems will double in power and then have enough computing power to run this kind of bloated crap I'm putting out without being as slow as a tired snail." It's as much true of the mainstream Linux distro makers as much as it is true of Microsoft and other proprietary software vendors.
Just for my workaday Linux distro, Red Hat 7.1. I for the life of me cannot understand why in heaven's name I need to install Kerberos to install the RPM package for CVS or LPRng. I don't have a Kerberized network and have no intention of setting such a creature up anytime in the near future, and likely it's true for most everyone. Or why I'm forced to install Japanese TTF fonts (xtt-fonts) just to get GhostScript up and running, or why printconf has to have a Kanji converter (nkf). I don't read Japanese, and I imagine the vast majority of the users of Red Hat's standard edition will never have any need to view, much less print, a Japanese-language document. The list of odd dependencies can go on and on ad nauseam, and there are many other signs of bloat. It's this kind of bloat that makes it impossible to run an up to date Linux distro on older hardware.
The other problem comes from hardware manufacturers, which is why unless Moore's Law comes to an end someday, this trend is going to keep going. And never mind us folks whose incomes cannot support a major hardware upgrade every 18 months. When a new technology appears, they stop making the old technology almost instantly. Can you still buy EDO SIMM's? Can you still buy a non-AGP video card? Well, unless you go to a surplus shop, probably not. Because of Moore's Law and its effect on the market, obsolete hardware has a way of becoming impractical or even impossible to maintain at some point, which is why everyone, even us in the third world who don't have a lot of disposable income and can't constantly support hardware upgrades, is eventually forced to upgrade.
While this project's aims are commendable, I wouldn't hold out too much hope for a universal adoption of its philosophy, not until Moore's Law comes to an end and the computer hardware market stabilizes as a result. Until then, I hope they remain true to the vision and not succumb to the temptations that have created the bloated monstrosities common nowadays.
Tell me about it. The authors (the so-called "Gang of Four") of the book Design Patterns: Elements of Reusable Object-Oriented Software drew on the work of the architect Christopher Alexander for their seminal work in object oriented design. You might even go so far as to say that the purest fusion of art and technology is really architecture in this more general sense.
...the PBS site goes down? It usually takes a week after he writes something new before I can see it. Is it because www.pbs.org is some dinky old server that takes a week to get up again after a slashdotting? Lately Slashdot's been linking to every article he writes, so of course, everyone's gonna try to look at it. This behavior has been consisten since I started reading Cringely's articles regularly last year.
An ancient verse in Open SOurce Lore...
Gee, so I guess that makes Miguel de Icaza Celebrimor, building his own runtime based on secrets given to him by the Dark Lord of Redmond, disguised as Annatar, Lord of Gifts. Maybe RMS is Elrond, watchful and distrustful of this mysterious being bearing secrets...
Well, coming from the point of view of traditional AI research it is truly radical. Call that the straitjacketed minds of crusty philosophers stuck in the ivory towers of academe, caught up in the biases of Western thought that seeks to divide, compartmentalize, and analyze the system of the world to understand it!
Brooks himself got these ideas from biology, a study so very far removed from the fields of computer science and electrical engineering that form the core of traditional AI research. It was only by stepping outside the bounds of traditionalist Western ideas about the compartmentalization of learning and knowledge that he brought these ideas forth.
I wonder what other ideas might come from a more integrated view of science, as opposed to the divisive approach Western science has taken.
Right. But then, a 'creature' with those kinds of 'sense organs' would be completely different and utterly alien to us. Because its experience of the world is utterly different, its emergent behavior would also be utterly different. In order to create a human-like creature, with human-like intelligence, the holy grail of AI, it would then necessarily have to have the sensory capabilities that a human would have. Either that, or you model the entire environment a human normally interacts with and allow your "artificially intelligent" being to interact with that simulated world, which is what traditional AI is trying to do. Of course, it's almost completely impossible to do that in its fullest generality... Very old arguments put forth by Hubert Dreyfus and Joseph Weizenbaum in the mid-1960's.
I wonder if he talks about Professor Rodney A. Brooks at MIT and his ideas about artificial intelligence, situatedness, and embodiment.
For Rod Brooks, "intelligence" cannot really be programmed into a system; it is rather an emergent property of systems as they interact with their environment. In The Matrix Morpheus says that the body cannot exist without the mind, but Brooks would rather say that the mind cannot exist without the body, because the body is the only way that the mind can have any experience of its environment. It's a radical idea. It answers the problems behind knowledge representation that have been argued by Hubert Dreyfus in 1965, where he stated that any representation of knowledge is incomplete without its connection to all other pieces of knowledge. The paradigm Brooks is presenting in his ideas about embodied intelligence is that explicit representation of knowledge is superfluous: let the world itself be its own best model, and let the artificially intelligent being formulate its own judgments about what the world is and what it means from its own experience of that world. Intelligence emerges from its interaction and experience of the world. If Brooks is correct, then true AI is absolutely inseperable from robotics.
The seminal paper where Brooks discusses this philosophy is "Intelligence Without Reason" and is available at his website which is linked above.
Any book on AI that does not discuss this other branch of AI philosophy is in my view hopelessly incomplete.