Pi is not really "Random"
on
Share The Pi!
·
· Score: 1
If you think about it, the digits of pi are not really a "random sequence" at all, at least according to Gregory J. Chiatin's theory of algorithmic information theory. The digits of Pi are of course compressible. You can write a computer program which is of finite size that will generate the digits of Pi, and that's definitely smaller than all the digits! The "randomness" only arises from our choice of base, actually. If you would use a factorial base representation (for instance) to write Pi, it wouldn't look very random...
"The model 100 series is completely data compatible with every computer on the face of the earth."
I suppose this old old machine which is older than many slashdotters still uses the standard RS-232 serial ports to do its interfacing to the outside world. It's one of the standards that has managed to endure, and it's probably only because this standard did endure that the TRS-80 laptops haven't died out. It's what allows you to use these Model 100's even in today's age. Without the ability to interface to the outside world no machine is worth beans.
I don't know, but I think maybe RS-232's days as a ubiquitous standard are numbered. I recently bought an IBM ThinkPad which doesn't have any RS-232 ports, only USB's, which have caused me a great amount of grief (and no small amount of money as well, I shelled out the equivalent of US$50 for a USB to serial converter, no small change out here in the Third World!) attempting to interface it with my Palm. Will this trend continue, I wonder?
Hmmm, let's see. There's the Russian Federation (lots of missiles, obviously). There's China (DF-5/CSS-4, range 13,000 km). There's India (Surya, range of 12,000 km). The other declared nuclear powers (Pakistan, France, UK, and Israel) have IRBM's only at the very most, according to publicly available information, and with one exception these are all allies of the United States. The three other powers which do have ICBM systems are unlikely to use them against the United States. The expense involved in an elaborate ICBM defense system is thus difficult to justify in this case. See the FAS Nuclear Forces Guide for more facts and figures regarding nuclear forces. They got informative pages on other weapons of mass destruction as well.
Jeez, these people are much too late. They want to stop the Chinese from getting DVD technology? I think they already had it long ago. If the PRC can steal U.S. Government nuclear secrets I think getting DVD technology trade secrets would be child's play for their equivalent of the KGB. There has to be some other reason, I think, or else these people in the U.S. and Japanese governments have either completely lost their mind or are under the influence of some controlled substance.
2437470 source lines of code for the Linux kernel. Doesn't that worry some people out there? We have a monolithic kernel almost two and a half million lines long. I think that by 2.6 the kernel is going to collapse under its own weight unless the designers decide to reorganize it in a fundamental way. Maybe it's time for a Linux-Hurd fusion project that will turn Linux into a true microkernel.
Let's see... A typical stereo sound card has 2 A/D-D/A channels sampling at 44 kHz at 16 bits per sample. So that gives us 44,000*16*2 = 1.4 Mbps. Not too bad I think, but most likely in practice you'll get only a small fraction of this, quite possibly comparable to 128k ISDN or less on the average. A SB16 can only do full duplex with either the A/D running at 16 bits/sample and the D/A running at 8 bits/sample and vice-versa. And one of them will have to be restricted to mono. Which means maximum upload bandwidth will probably be less than a quarter of the download bandwidth, which is not too bad as for the average user, this is made up primarily of HTTP/GET requests:).
Well, the article says that he was actually able to build a neutron gun to transform Thorium 232 into Uranium 233. If he decided to get a little more, he could have a small sphere of sufficiently enriched U-233, and along with the beryllium he got, and some high explosive arranged in the proper way, would be sufficient to create a crude nuclear device. Makes the term "nuclear terrorism" sound a lot more ominous... A 15-year-old kid was able to build a makeshift nuclear reactor in his backyard with common household objects. How hard can it be for determined terrorist organizations to build a nuclear device? Isn't even necessary to build a bomb, all they need is the neutron gun and some cobalt (which is not even a restricted material). In a few days, you'd have a chunk of Cobalt-60 more radioactive than Chernobyl, and it'd just be a matter of grinding it down and dispersing it in a suitable area... Enough to poison an area to all kinds of life for many many years.
The ccTLD for my own country (.ph) is itself being torn by extremely harsh in-fighting. There have been calls for redelegation, and violent flamewars on a yahoogroups mailing list (dotph-issues) created to address this issue. I doubt if Joel Disini, the head of DotPH Inc. will bother too much with this issue. He's already getting loads of flak from the local Internet scene because he is perceived to have abused his authority as domain administrator, by allegedly giving preferential treatment to the ISP he owns, among other things. If the DNSO and ICANN are also embroiled in their own internal conflict, it may be unlikely that our issues with our own ccTLD administrator will be resolved as expeditiously as we hope.
I just wonder if it would be technologically feasible and practical to actually create a peer to peer application that would provide all the functions of the current Domain Name Service. One way to get rid of ICANN forever, and eliminate the weak link that binds the Internet (no pun intended). If the practical functions of the domain servers they own and control are taken away, well, that would be the end of them, I think. That will mean the beginning of the end for all dreams of central control over the Internet.
Well, you're using commercial software, right? You didn't just pay for a chunk of bits and bytes when you got it. Get your money's worth and call them to ask what changes the program will make to your system configuration. I think that's the easiest and simplest way.
We're going back to moving parts here. At least a battery-based UPS would have minimal moving parts, probably a few relays for overcurrent protection and a transformer with a moving center tap to regulate voltage in the traditional UPS. With a big moving part like a flywheel, we have got a fair bit of noise here. The specs say less than 55 dB, which is roughly the amount of noise produced by a typical conversation. And the efficiency of the system is somewhat pathetic. The system can store up to 2 kWh of energy. It takes three hours for it to store up this much energy with an input power of 2.5 kW, meaning you input 7.5 kWh of energy to it. The remaining 5.5 kWh of energy was wasted. Wasting so much energy is not the way to be environmentally friendly! True, a traditional battery based UPS would cause environmental pollution when you tried to dispose of it, but 5.5 kWh of wasted energy every time you had to recharge the thing would likely produce even more pollution in total. True, the system may have many potential advantages, especially for use in harsh environments that would ruin most battery-based systems, but to call it environmentally friendly is an absolutely laughable claim. It is probably even less environmentally friendly than the average battery UPS, it just produces a different kind of pollution (by wasting much of the energy provided to it from the power generation system). And one more thing: the thing is immense, it weighs a total of 1050 lbs!
We don't wish to be antisocial. It's just that society has put a stigma on us JUST FOR BEING DIFFERENT. I suppose everyone here has read JonKatz's series on the Hellmouth. They are the ones who ostracize us to begin with, and we wind up ostracizing them back. The vicious cycle needs to be broken.
I believe that New Zealand is one place. It's even said (on OpenDVD.org) that there seems to be a clause in the WTO treaties that makes these sorts of things illegal. I'm not sure exactly what New Zealand does, perhaps they made it illegal to import or sell any DVD player that honors region codes.
It's not hard to see why the ACCC would find this in violation of Australian consumer protection law. When you look at DVD region coding carefully, you find it's nothing more or less than an attempt by the MPAA to perform price fixing on a global scale. Why not all nations are up in arms against this yet is testimony to the power of these movie studios.
Java did not really introduce any major innovations in programming language design, at least not anything that was not already present in many other programming languages. Java is actually C++ on a diet, which is also bytecode interpreted to provide what Sun thought would be platform-independence, but in reality it made very slow code. Try using the Java and C++ versions of the Xerces and Xalan XML and XSLT parsers available from the Apache XML project to do the same thing, such as convert even a small DocBook document into HTML with an appropriate set of XSL stylesheets. The difference is striking. (in my case, I have no choice, because there is no non-Java, Free [speech] XSL Formatting objects converter with functionality comparable to FOP, which is not only in Java but depends on the Java versions of Xerces and Xalan)
As for what one feature I'd like to see for new programming languages is CLEAN DESIGN. I like the way R^4RS Scheme document put it: "Programming languages should not be designed by piling feature on feature, but by removing the weaknesses and restrictions that make additional features seem necessary." It seems that most languages that have existed primarily as open source projects and even many standardized languages wind up doing precisely the opposite, piling features that seem necessary but could be better be served by extending what features are already there to make the features unnecessary. What I'd like to see is an object-oriented or imperative language that tries to emulate Scheme's philosophy: the attainment of perfection by finding the point where there is nothing more that can be taken away.
Inter.net was created by joining together all of PSINet's consumer Internet and portal businesses and is the only global consumer Internet Service Provider. While PSINet will retain a minority ownership position in Inter.net Global, there will be no future financial obligation to Inter.net Global and no seats on the Board of Directors.
I guess this means that all former small business subscribers to PSInet may actually get folded to Inter.Net. Since I work for the Philippine subsidiary, I have no idea how this might work in the US or if this might actually happen.
Is that it's somewhat one-sided. We all know Tim O'Reilly, Larry Lessig, John Perry Barlow, Dan Gillmore, and Clay Shirky all believe in more or less the same thing: that information is not a product or property to be traded and sold, but a service to be provided, and that is the crux of this whole question. Now, if they had gotten one of the people whom Eben Moglen calls 'econodwarfs' or 'IPdroids' in his seminal essay: Anarchism Triumphant: Free Software and the Death of Copyright the debate might have been much more interesting. Maybe if they got Jack Valenti or those other partisans of strong "intellectual property" rights to participate...
Considering that an enterprise relational DB is "boring" work compared to an OS, I don't think that either MySQL or PostgreSQL will ever grow as quickly as either Linux or Apache have. Making a full relational database which passes the ACID test is a non-trivial task, and while PostGreSQL of the two free RDBMS's has made the most progress towards this goal, I don't think it's capable of handling the terabyte-sized datasets that Oracle routinely does, and it will be a long time before anyone even thinks of even trying to do it. Why do you suppose there are only two serious contenders for a Free Software enterprise RDBMS (and up until recently, one of them didn't even fit that description entirely!)? An RDBMS is not as sexy as an OS or a desktop environment, not the sort of thing that would capture the imagination of the hackerly community. Finally, an industrial strength RDBMS like Oracle doesn't fill the needs of the hackerly community, but those of the business community. A site like slashdot certainly needs an RDBMS, but it doesn't manage terabytes of data, and neither will it be a major catastrophe if the database fails because either A, C, I, or D was not followed (it may cause inconvenience at the very most, but probably nobody is going to die or lose money as a result). Frankly, I think Larry Ellison is safe for the moment. Nothing that the Free Software World has to offer can really match Oracle at the moment. The same cannot be said of Bill Gates, though.
In my opinion, one of the most awful features of Python as a language is the fact that it thinks of whitespace as a highly significant feature in programs, so much so that you always have to properly indent all blocks of code that are under a particular control structure (such as if statements or for loops), instead of having a 'begin'-'end' token pair to do the job, such as what we have in C, Perl, Java, Pascal, and almost every other block structured programming language I know about. It takes away the programmer's freedom to style his or her code, forcing them to conform to somebody's idea about how programs ought to visually look like. I've heard of "bondage and discipline" languages, but this is arch-B&D... not even Pascal is so anal!
You know, this reminds me a lot about the Yamashita treasure that was supposed to have been hidden in my country. I guess Ferdinand Marcos found it and it's now part of the stash the late dictator's family has in Switzerland. Most likely given the number of military dictators who've held rule in Thailand it's probably long gone by now.
Most so-called Heisenbugs are really race conditions, and their behavior looks non-deterministic only because the interactions involved with context switches and volatile variables can become extremely complicated. The best way to remove heisenbugs is to program defensively, taking care to prevent possible races from occuring. Examine all the possible places shared data structures can be changed and make sure that they are properly protected by mutexes. Ensure that the mutexes are updated in the proper order to avoid possible deadlocks. Printf debugging works better than any other method here I find.
Already, a lot of truly historic correspondence from the dawn of the Digital Age is possibly gone forever. As another Slashdot article some time ago pointed out, we can read Galileo's correspondence on astronomy in the 16th century but we can't read Marvin Minsky's correspondence on AI in the 20th century. The pace at which stoarage technology is changing is moving so rapidly, and these concerns about "digital rights management" that CPRM and the DVD-CCA are supposed to address will make it even more difficult. These "Digital Rights" basically include the "right to allow your content to vanish in the mists of time".
Considering how complex Perl is, I doubt that someone would be able to make a compiler for it that would be under 2 megs. Add that to the overhead it would need to interpret the bytecode, any conceivable complete implementation would be useless on any Palm released to date. Add to that the fact that the Motorola Dragonball processor that lives at the heart of all Palm Computing Devices is slower than any computer that ever possessed a Perl system to date, and it's even more worthless.
Much more useful would be a Perl system that accepted only a small subset of the full Perl language, but deciding what subset to use is the big trick. Getting rid of associative arrays and the OOP cruft bolted on Perl might be useful, but it's probably insufficient to simplify the language enough.
That is, unless someone finds an innovative way of structuring the compiler/interpreter system of Perl, just like the 8-bit JVM which ran on/. a while back, but it's too much to be hoping for, I believe.
Slashdot Mirror
If you think about it, the digits of pi are not really a "random sequence" at all, at least according to Gregory J. Chiatin's theory of algorithmic information theory. The digits of Pi are of course compressible. You can write a computer program which is of finite size that will generate the digits of Pi, and that's definitely smaller than all the digits! The "randomness" only arises from our choice of base, actually. If you would use a factorial base representation (for instance) to write Pi, it wouldn't look very random...
I suppose this old old machine which is older than many slashdotters still uses the standard RS-232 serial ports to do its interfacing to the outside world. It's one of the standards that has managed to endure, and it's probably only because this standard did endure that the TRS-80 laptops haven't died out. It's what allows you to use these Model 100's even in today's age. Without the ability to interface to the outside world no machine is worth beans.
I don't know, but I think maybe RS-232's days as a ubiquitous standard are numbered. I recently bought an IBM ThinkPad which doesn't have any RS-232 ports, only USB's, which have caused me a great amount of grief (and no small amount of money as well, I shelled out the equivalent of US$50 for a USB to serial converter, no small change out here in the Third World!) attempting to interface it with my Palm. Will this trend continue, I wonder?
Hmmm, let's see. There's the Russian Federation (lots of missiles, obviously). There's China (DF-5/CSS-4, range 13,000 km). There's India (Surya, range of 12,000 km). The other declared nuclear powers (Pakistan, France, UK, and Israel) have IRBM's only at the very most, according to publicly available information, and with one exception these are all allies of the United States. The three other powers which do have ICBM systems are unlikely to use them against the United States. The expense involved in an elaborate ICBM defense system is thus difficult to justify in this case. See the FAS Nuclear Forces Guide for more facts and figures regarding nuclear forces. They got informative pages on other weapons of mass destruction as well.
Jeez, these people are much too late. They want to stop the Chinese from getting DVD technology? I think they already had it long ago. If the PRC can steal U.S. Government nuclear secrets I think getting DVD technology trade secrets would be child's play for their equivalent of the KGB. There has to be some other reason, I think, or else these people in the U.S. and Japanese governments have either completely lost their mind or are under the influence of some controlled substance.
2437470 source lines of code for the Linux kernel. Doesn't that worry some people out there? We have a monolithic kernel almost two and a half million lines long. I think that by 2.6 the kernel is going to collapse under its own weight unless the designers decide to reorganize it in a fundamental way. Maybe it's time for a Linux-Hurd fusion project that will turn Linux into a true microkernel.
Let's see... A typical stereo sound card has 2 A/D-D/A channels sampling at 44 kHz at 16 bits per sample. So that gives us 44,000*16*2 = 1.4 Mbps. Not too bad I think, but most likely in practice you'll get only a small fraction of this, quite possibly comparable to 128k ISDN or less on the average. A SB16 can only do full duplex with either the A/D running at 16 bits/sample and the D/A running at 8 bits/sample and vice-versa. And one of them will have to be restricted to mono. Which means maximum upload bandwidth will probably be less than a quarter of the download bandwidth, which is not too bad as for the average user, this is made up primarily of HTTP/GET requests :).
Well, the article says that he was actually able to build a neutron gun to transform Thorium 232 into Uranium 233. If he decided to get a little more, he could have a small sphere of sufficiently enriched U-233, and along with the beryllium he got, and some high explosive arranged in the proper way, would be sufficient to create a crude nuclear device. Makes the term "nuclear terrorism" sound a lot more ominous... A 15-year-old kid was able to build a makeshift nuclear reactor in his backyard with common household objects. How hard can it be for determined terrorist organizations to build a nuclear device? Isn't even necessary to build a bomb, all they need is the neutron gun and some cobalt (which is not even a restricted material). In a few days, you'd have a chunk of Cobalt-60 more radioactive than Chernobyl, and it'd just be a matter of grinding it down and dispersing it in a suitable area... Enough to poison an area to all kinds of life for many many years.
The ccTLD for my own country (.ph) is itself being torn by extremely harsh in-fighting. There have been calls for redelegation, and violent flamewars on a yahoogroups mailing list (dotph-issues) created to address this issue. I doubt if Joel Disini, the head of DotPH Inc. will bother too much with this issue. He's already getting loads of flak from the local Internet scene because he is perceived to have abused his authority as domain administrator, by allegedly giving preferential treatment to the ISP he owns, among other things. If the DNSO and ICANN are also embroiled in their own internal conflict, it may be unlikely that our issues with our own ccTLD administrator will be resolved as expeditiously as we hope.
I just wonder if it would be technologically feasible and practical to actually create a peer to peer application that would provide all the functions of the current Domain Name Service. One way to get rid of ICANN forever, and eliminate the weak link that binds the Internet (no pun intended). If the practical functions of the domain servers they own and control are taken away, well, that would be the end of them, I think. That will mean the beginning of the end for all dreams of central control over the Internet.
Well, you're using commercial software, right? You didn't just pay for a chunk of bits and bytes when you got it. Get your money's worth and call them to ask what changes the program will make to your system configuration. I think that's the easiest and simplest way.
We're going back to moving parts here. At least a battery-based UPS would have minimal moving parts, probably a few relays for overcurrent protection and a transformer with a moving center tap to regulate voltage in the traditional UPS. With a big moving part like a flywheel, we have got a fair bit of noise here. The specs say less than 55 dB, which is roughly the amount of noise produced by a typical conversation. And the efficiency of the system is somewhat pathetic. The system can store up to 2 kWh of energy. It takes three hours for it to store up this much energy with an input power of 2.5 kW, meaning you input 7.5 kWh of energy to it. The remaining 5.5 kWh of energy was wasted. Wasting so much energy is not the way to be environmentally friendly! True, a traditional battery based UPS would cause environmental pollution when you tried to dispose of it, but 5.5 kWh of wasted energy every time you had to recharge the thing would likely produce even more pollution in total. True, the system may have many potential advantages, especially for use in harsh environments that would ruin most battery-based systems, but to call it environmentally friendly is an absolutely laughable claim. It is probably even less environmentally friendly than the average battery UPS, it just produces a different kind of pollution (by wasting much of the energy provided to it from the power generation system). And one more thing: the thing is immense, it weighs a total of 1050 lbs!
We don't wish to be antisocial. It's just that society has put a stigma on us JUST FOR BEING DIFFERENT. I suppose everyone here has read JonKatz's series on the Hellmouth. They are the ones who ostracize us to begin with, and we wind up ostracizing them back. The vicious cycle needs to be broken.
I believe that New Zealand is one place. It's even said (on OpenDVD.org) that there seems to be a clause in the WTO treaties that makes these sorts of things illegal. I'm not sure exactly what New Zealand does, perhaps they made it illegal to import or sell any DVD player that honors region codes.
It's not hard to see why the ACCC would find this in violation of Australian consumer protection law. When you look at DVD region coding carefully, you find it's nothing more or less than an attempt by the MPAA to perform price fixing on a global scale. Why not all nations are up in arms against this yet is testimony to the power of these movie studios.
Well, isn't developing for Linux a community service? ;)
Pot...Kettle...Black.
Java did not really introduce any major innovations in programming language design, at least not anything that was not already present in many other programming languages. Java is actually C++ on a diet, which is also bytecode interpreted to provide what Sun thought would be platform-independence, but in reality it made very slow code. Try using the Java and C++ versions of the Xerces and Xalan XML and XSLT parsers available from the Apache XML project to do the same thing, such as convert even a small DocBook document into HTML with an appropriate set of XSL stylesheets. The difference is striking. (in my case, I have no choice, because there is no non-Java, Free [speech] XSL Formatting objects converter with functionality comparable to FOP, which is not only in Java but depends on the Java versions of Xerces and Xalan)
As for what one feature I'd like to see for new programming languages is CLEAN DESIGN. I like the way R^4RS Scheme document put it: "Programming languages should not be designed by piling feature on feature, but by removing the weaknesses and restrictions that make additional features seem necessary." It seems that most languages that have existed primarily as open source projects and even many standardized languages wind up doing precisely the opposite, piling features that seem necessary but could be better be served by extending what features are already there to make the features unnecessary. What I'd like to see is an object-oriented or imperative language that tries to emulate Scheme's philosophy: the attainment of perfection by finding the point where there is nothing more that can be taken away.
I found this quote on my company's US website:
The whole link is here.
I guess this means that all former small business subscribers to PSInet may actually get folded to Inter.Net. Since I work for the Philippine subsidiary, I have no idea how this might work in the US or if this might actually happen.
Is that it's somewhat one-sided. We all know Tim O'Reilly, Larry Lessig, John Perry Barlow, Dan Gillmore, and Clay Shirky all believe in more or less the same thing: that information is not a product or property to be traded and sold, but a service to be provided, and that is the crux of this whole question. Now, if they had gotten one of the people whom Eben Moglen calls 'econodwarfs' or 'IPdroids' in his seminal essay: Anarchism Triumphant: Free Software and the Death of Copyright the debate might have been much more interesting. Maybe if they got Jack Valenti or those other partisans of strong "intellectual property" rights to participate...
Considering that an enterprise relational DB is "boring" work compared to an OS, I don't think that either MySQL or PostgreSQL will ever grow as quickly as either Linux or Apache have. Making a full relational database which passes the ACID test is a non-trivial task, and while PostGreSQL of the two free RDBMS's has made the most progress towards this goal, I don't think it's capable of handling the terabyte-sized datasets that Oracle routinely does, and it will be a long time before anyone even thinks of even trying to do it. Why do you suppose there are only two serious contenders for a Free Software enterprise RDBMS (and up until recently, one of them didn't even fit that description entirely!)? An RDBMS is not as sexy as an OS or a desktop environment, not the sort of thing that would capture the imagination of the hackerly community. Finally, an industrial strength RDBMS like Oracle doesn't fill the needs of the hackerly community, but those of the business community. A site like slashdot certainly needs an RDBMS, but it doesn't manage terabytes of data, and neither will it be a major catastrophe if the database fails because either A, C, I, or D was not followed (it may cause inconvenience at the very most, but probably nobody is going to die or lose money as a result). Frankly, I think Larry Ellison is safe for the moment. Nothing that the Free Software World has to offer can really match Oracle at the moment. The same cannot be said of Bill Gates, though.
In my opinion, one of the most awful features of Python as a language is the fact that it thinks of whitespace as a highly significant feature in programs, so much so that you always have to properly indent all blocks of code that are under a particular control structure (such as if statements or for loops), instead of having a 'begin'-'end' token pair to do the job, such as what we have in C, Perl, Java, Pascal, and almost every other block structured programming language I know about. It takes away the programmer's freedom to style his or her code, forcing them to conform to somebody's idea about how programs ought to visually look like. I've heard of "bondage and discipline" languages, but this is arch-B&D... not even Pascal is so anal!
You know, this reminds me a lot about the Yamashita treasure that was supposed to have been hidden in my country. I guess Ferdinand Marcos found it and it's now part of the stash the late dictator's family has in Switzerland. Most likely given the number of military dictators who've held rule in Thailand it's probably long gone by now.
Most so-called Heisenbugs are really race conditions, and their behavior looks non-deterministic only because the interactions involved with context switches and volatile variables can become extremely complicated. The best way to remove heisenbugs is to program defensively, taking care to prevent possible races from occuring. Examine all the possible places shared data structures can be changed and make sure that they are properly protected by mutexes. Ensure that the mutexes are updated in the proper order to avoid possible deadlocks. Printf debugging works better than any other method here I find.
Already, a lot of truly historic correspondence from the dawn of the Digital Age is possibly gone forever. As another Slashdot article some time ago pointed out, we can read Galileo's correspondence on astronomy in the 16th century but we can't read Marvin Minsky's correspondence on AI in the 20th century. The pace at which stoarage technology is changing is moving so rapidly, and these concerns about "digital rights management" that CPRM and the DVD-CCA are supposed to address will make it even more difficult. These "Digital Rights" basically include the "right to allow your content to vanish in the mists of time".
"Now Mr. Cringely... what did you say 2+2 was again?"
"Four"
"And if Big Brother Bill says 2+2 is five, or one billion, what is 2+2?"
"Four--" His answer vanished in a scream of pain.
SLOWNESS IS SPEED
CRASHING IS STABILITY
IGNORANCE IS STRENGTH
--somewhere at Redmond
Considering how complex Perl is, I doubt that someone would be able to make a compiler for it that would be under 2 megs. Add that to the overhead it would need to interpret the bytecode, any conceivable complete implementation would be useless on any Palm released to date. Add to that the fact that the Motorola Dragonball processor that lives at the heart of all Palm Computing Devices is slower than any computer that ever possessed a Perl system to date, and it's even more worthless.
Much more useful would be a Perl system that accepted only a small subset of the full Perl language, but deciding what subset to use is the big trick. Getting rid of associative arrays and the OOP cruft bolted on Perl might be useful, but it's probably insufficient to simplify the language enough.
That is, unless someone finds an innovative way of structuring the compiler/interpreter system of Perl, just like the 8-bit JVM which ran on /. a while back, but it's too much to be hoping for, I believe.