I'll give you an excellent reason for them suing Sony and not IBM. IBM is an 8000 lb. gorilla that has an army of vampiric lawyers so skillful they can not only draw blood from a stone, they can even draw blood from SCO. Sony's lawyers, on the other hand, have trouble with lawsuits against preteen fileswappers and computerless grannies. If you want to leech money, it's perfectly obvious which one you're more likely to leech successfully from.
And, no, the patent is not valid. Such devices have existed for thirty years, albeit mostly in research labs and Universities. Prior art is prior art, however, and Sony needs only one example that is close enough to invalidate the patent.
It is just possible that this lawsuit is funded by Microsoft or some other console manufacturer, in an effort to throw legalistic mud at Sony's console and thus boost sales of their own. Same way SCO tried to boost their sales by throwing legalistic mud at Linux. I would not rule that out just yet, as this may well impact sales. People aren't going to want to buy a console they can't do anything with and will have to destroy a few weeks later under court orders.
Oh, there's no shortage of choices. There'd be even more, if Inmos hadn't been sold off and their chips relegated to video recorders. Lambda calculus is fine, pi-calculus is also good. ISO/IEC 13568:2002 would be most desirable. In fact, anything amenable to formal quality control would be good.
If a desktop-ish OS were to be used, LynxOS (a Linux offshoot that has some respectable avionics certification) would not be a bad choice. I'd be a little concerned about vanilla Linux, but if you stuck to the better-tested components and ripped the rest out, then really worked on validating what was left to a high standard, I'd consider it a doable option.
Haiku might also be workable, with some work, as it's designed to support a very large number of independent activities - exactly the kind of environment you'd face in a car.
In the Valley of the Kings, there lies a Great Glass Pyramid. If you'd rather something that wasn't being used, the tower hasn't housed an accelerator for over fifteen years. Problem solved.
Can you give me something at least a little challenging? A canal that runs uphill, a viaduct that can span a couple of cities in an earthquake zone with a 100+ year warranty, a Michelangelo-style work on the inside of a glacier - you know, something hard. Something that might be a bit more challenging than merely keeping something safe from the elements. Unfortunately, those were all done in that general area, so you'll need to pick something else.
Last I heard, there were only three being exhibited, but that doesn't mean there aren't more of them. And "left to rot" is correct - they've all suffered severe damage from corrosion, weathering, assorted plants and animals, etc. The one on loan from the Smithsonian was almost beyond repair, according to some reports, with massive amounts of replacement material being required. You treat anything on loan from me like that and you'd be making a close-up inspection of Mount St. Helens' current mountain-building activity. But because it's merely a stupendously rare internationally significant historic artifact marking the entry into genuine manned space travel, who gives a damn if it's destroyed by insects, squirrels and pollution? Apparently not the center it is parked in.
In an anonymous system, it is (by definition) impossible to match up a single paper vote with a single electronic vote - always assuming the electronic votes are kept separate at all.
The ONLY workable system is to keep all individual votes separate in both forms and re-tally them independently. It makes no difference then whether the paper copy is identical to the corresponding electronic copy, if the grand totals for each add up to the same value. In which case, why would it matter if the sequences were not identical, provided the totals were?
However, it would be easy enough to have even that possibility checked for. Have each vote marked with a 4096-bit string that is either totally random or some sort of extremely strong cryptographic hash of the vote plus time plus other data. All you need is values that are unique and preserve anonymity. (They can't therefore be sequential or trivially derived from things that are knowable.) This would be included in both the electronic copy and the paper copy. If the extra checks are needed, simply look for unique vote IDs that exist in one system but not in the other, or for unique vote IDs where the two versions of the vote are not the same.
Is there anything else that can be done? Sure. Public key encryption can be used to digitally sign things, and the public and private keys should be kept apart. Simply put on each voter registration card an electronic tag with a public key. Two copies of the private key are made. One is at the central voting office and one is kept with independent electoral observers. When a vote is cast, the vote + unique vote ID is digitally signed. When a discrepancy is found between what is electronically cast and what is on paper, you check the votes with the suspect unique vote IDs. For each of those votes, exactly one private key (owner unknown) should match, and vice versa. Just run through the keys until one matches, then mark that key as used. If a key needs to be used more than once, then you have evidence of attempted ballot box stuffing and you can remove all votes signed with the same key.
Ok, anything else? Sure. When a vote is electronically cast, one copy goes to the central voting office (votes should NEVER be counted at the local voting center) and one copy should go to the independent observers. Both copies should be encrypted using public key encryption per voting center - not only to allow counting by center but also to validate that no center is getting more votes than voters and to prevent man-in-the-middle attacks. The independent observers should have statutory powers to challenge an election result if they detect inconsistencies, errors in procedure, flawed tallies and failures within the official body to deal with errors.
However, we now hit the hardest problem of all. As Plato noted in his "Republic", voter stupidity is the greatest flaw in democracy. When you look at the amount spent on education - just under one point three billion - it seems like a lot. When you examine how much that works out to, it's about $50 per person per year. Dunno about you, but $50 isn't going to do much educating. I doubt it'll even cover getting people familiar with a new voting system. If the only way democracy can work is through an enlightened populace, then no voting system will cure the flaws present in America.
Sure, but then I have always been a little.... odd.... The doctors say they can treat most of it, but my insurance won't cover the surgery needed to stop turning into a bat.
I disagree. A zombie army will come in handy when fighting patent battles. Lawyers can't suck the lifeblood from a zombie, because it doesn't have any.
Most packet inspectors (such as Network Observer) are packet class only. Converged Access does a more sophisticated packet inspector, but even that only drills down to the specific subtype of packet for a given application, and of course only those applications they have the specifications for, or reverse-engineered. I know of no full-payload inspectors and doubt they even exist. Remember that packets cannot be guaranteed to travel on identical paths - the Internet is not a spanning tree - and that packets can fragment when there is an MTU change. Anyone sending a jumbo packet is guaranteed to see packet fragmentation, for example.
A full reassembly by sniffing would also need to drop retransmitted packets and support all common encapsulation techniques. You're also talking about a LOT of storage and absolutely no way to sensibly organize the volume of data collected. That's the problem with data saturation - there are no database or data processing techniques capable of handling it. I was talking to one of the top Ingres software/network gurus at OSCON yesterday - apparently even just the total information awareness project is staggering under the sheer weight of information that no system yet designed can handle. If the data is unsearchable, unsortable and unprocessable, then to all practical intents and purposes, it doesn't exist.
AMD also needs the One Laptop project to buy chips - they're potentially a gigantic customer, far bigger than any domestic slice of the pie Dell might have - and OLPC is also Open Source. I don't know exactly what graphics OLPC uses in the current version for graphics, but you can bet your left sock that AMD is going to want to supply as large a fraction of the components as possible.
Dell doesn't have fans on Slashdot, but I doubt they're doing that badly in Slashdot League Division 1, and they seem to be improving on the Good Idea/Bad Idea ratio. Of course, that could just be a phase they're going through too.
Political organizations are supposed to define policies, management should be left to managers, and actual work should be left to the workforce.
Sun Tzu's classic document "Art of War" makes it very clear that you should NEVER have a politician actually commanding the armed services. The same logic goes for all other departments. Politicians are very good for looking at the big picture (well, in theory) in a way that specialists in individual fields cannot. That makes them good for determining priorities, allocating resources, setting long-term objectives, etc. But once they have issued those decisions, the rest should be entirely left to those who are competent in the field -- with one exception. Governing entitles politicians to penalize those who violate the rules necessary for a coherent organization.
The modern idea that politicians should be in control is a bastardization of the entire concept of a democracy or republic. Plato's Republic is a little dated, but does explain the difference between a ruling class and a governing class. This is an important distinction and one that many have apparently forgotten. Rulers rule. They impose. That is their nature, that is their job. If that is how you see American politics, then you are saying America has an elected monarch. (I believe the archaic term is Bretwalda, and yes elected kings have existed throughout history.) Governors govern. If the populace is the clay and the civil service are the artists, the government is nothing more than an art critic sponsoring the latest work. Nothing more.
Now, personally I don't believe that quality government exists. Here, there, or anywhere. I also generally believe that most existing Governments in the world are indeed elected monarchies... with the rest being hereditary monarchies, dictatorships and fiefdoms of various sorts. Despite the roots of constitutional law being over 5,000 years old, the notions of democracy reaching back over 2,500 years and the concept of politics as a science being studied and researched for many centuries, I can recall no time in history or in the modern world where anyone has actually applied any of these ideas.
To me, the question boils down to this. If everyone in America treats the Federal Government like a kingdom and the States like princedoms (yes, the term does exist), why not cut to the chase and cut costs at the same time by declaring it such? If people truly, honestly, believe that's what they have, then what are they going to miss by making it official? If, however, you believe that the Government is truly restricted to governing and nothing else, then you not only should imagine the government spending tax money without controlling how it is spent, you should require it.
I would agree that the vanilla kernel does indeed fail on the desktop. I would argue that X' single-threaded model is actually a worse culprit, however, as it makes it much harder to efficiently (or, indeed, selectively) utilize resources on modern systems. That it is a reference implementation and never intended for production desktops doesn't help.
Slashdot has thrashed X many a time, so I'll focus on the kernel's hangups. To the best of my knowledge, the PPS patches bit-rotted and nobody wrote a new implementation. This limits the timer resolution Linux can handle. The kernel burns cycles on polling. The kernel context switch and userspace-to-kernelspace copying are damn expensive, which matters a lot if you're making lots of calls to the sound and graphics layers. Anything that could reduce that would help. (Intelligent graphics cards are obviously faster, but most desktops don't have GPUs, let alone GPUs of the kind of power you'd want.)
Well, technically quangos are committees of totally unqualified individuals who are appointed on idealogical grounds, but for the sake of argument I'll assume they all have multiple personality disorder and can't be distinguished from a committee even as an individual.
The current President's practices are, admittedly, much more grievous than many of the Thatcher-era or Blair-era quangos - generally they didn't create such groups with the power to seek (or bestow) death sentences. Well, apart from a few dodgy shoot-to-kill types, who were investigated. Well, would have been, but Stalker got kicked off the job for political reasons. However, most of the sordid scandals have involved relatively petty financial irregularities. Both Thatcher and Blair sold honors, for example, and both "fixed" illegal arms contracts.
The problem being declared is fictional, the problem being solved (the stupidity of marketroids and execs) is not. And, yeah, I've no objection to spending a few minutes on a shell script if it means a sales guy is taken to the vets and fixed.
The appropriate professional experience for a President - or any politician - is as a liar, con-artist or conjurer. I can't think of any political types - in America or Europe - who failed to meet these requirements other than the former BBC war correspondent Martin Bell (who later revealed that British parliament was far more hazardous to human health than any war zone).
Look. It's quite simple. You have two moderately high-power machines, one running a copy of Slash and the various sections, but only having dummy articles, a dummy CSS and no images. It should also have all non-essential processes turned off and otherwise be configured for a massive thrashing. The second is directly connected to the first, using local gigabit ethernet. The second machine is set up to spoof a huge number of installs of Alexa, each on an independent account and IP address on this internal network. Again, the second machine is configured for raw performance. We don't care about quality, reliability, stability or usability. We care about building a fictional number of hits.
As far as the reporting mechanism is concerned, it is incapable of distinguishing between slashdot.org on the Internet and an identically-named, identically-numbered machine on a private network. As the fictional users are randomly changing, along with any registration info, from a pool that was collected earlier, there would be no possible way for the stat collecting firm to identify which users are real (on the real Slashdot) and which are fake (on the fake Slashdot).
What's more, because these are done on spare boxes, there is no overhead placed on the real site and the real servers are not burdened with pointless traffic.
Once you get an idea for how much fake isolated hits you can get, you then sell a hit-booster kit to other companies. They get a similar setup (fake server with real name, and a stress-testing platform with fake Alexa users) at so many dollars per fake increment on their stats. It should be a good money-spinner, if the unit increments are large enough.
Interestingly, partly from the same source. Only I bothered to follow the links to the updated figures.
Apparently you didn't notice estimates for 2006, which the National Autistic Society verifies as being 1:100 - somewhat higher than your 36:10000. In fact, how the hell did you get that number from the NAS, when the NAS' own figures (Estimated prevalence rate in the UK, given in the paper linked to above) are stated as being current and a replacement for the 1997 figures you cite.
However, you also managed to miss this little gem from the Autism Research Center at Cambridge University.
Sorry if I sound a little annoyed, but if you're going to question my figures, you need to do better than decade-old discredited reports. I don't mind people challenging my views - I enjoy the intellectual challenge - but I do object to put-downs that are out-of-date, based on flawed methods and demonstrate a lack of site-searching skills. Please! Even in D&D, I'm at a high enough level that the DM has to give the monsters PhDs in physics and advanced library science. (Ever been attacked by a goblin wielding a +20 Shakespere of Doom?)
I am professionally diagnosed with Asperger's, so have no trouble with stating so. The incidence of diagnosed Asperger's in the UK is now reportedly 1:56, which is three times the incidence rate in the US. It is unclear if this is because the UK has a superior mental health care system for diagnosing such conditions (that was not divulged in Sicko), whether the UK has been so attractive to eccentrics and odd-balls over the centuries that it is simply genetically more open to it, or whether the US' obsession with normalicy is artificially distorting the figures.
(For the record, I was diagnosed in the US, by two independent doctors.)
These days, diagnosis by a doctor based on external symptoms is wholly unnecessary - in theory. fMRI scans will reveal Asperger's and all other autistic spectrum disorders very nicely with a very high level of accuracy. The problem with such tests is not accuracy but expense and difficulty in finding a place equipped to carry them out. Neurologists who are sufficiently far up on such diagnostic methods, given that they've only circulated from conferences in the past couple of years, are extremely rare. So whilst a symptom-based diagnosis is not required, a mechanistic diagnosis is unlikely to be in anyone's near-term future.
(However, if Slashdotters do encounter a research group wanting to verify the published results, I'd say go for it. This would give you as close to definitive proof as you can ever hope for in medical science.)
Asperger's is problematic in that there is no cure and no therapy for most of it. It is also genetically as harmful as the gene that produces Manx cats. The gene for Manx cats is additive, so eventually not only does the tail go, but so does a large chunk of the spinal cord. The gene for the Autistic Spectrum seems to be similar - if your parents both are on the Autistic Spectrum, you will be further down the spectrum. Eventually, you have nothing but mentally deformed cabbages. Asperger's can be beneficial and should be utilized as such. Play to your strengths, not your weaknesses. However, it is also a very dangerous genetic trait and should not be trusted too much. Like a Ring of Power, over-use will lead to betrayal.
Trust me. If someone were to produce a clone of the latest Core2Duo with only one enhancement - zeroing a cache in half the time - and just enough cash to actually get a few chips circulating, Intel WILL reduce the time for that process by a few microseconds more by the end of the week. Same as they did with Transmeta and the power saving.
Nobody even needs to actually mass-produce the buggers - Transmeta's first batch didn't even work right and had to be recalled - they just need to scare Intel into thinking they could. So long as Intel believes they are the best, they will do nothing. The second they think they're not (even if it's just a skillful illusion) is the second they'll modernize the flawed part of their design.
Oh, eventually they'll catch on, as these mysterious competitors never actually appear, but if it's good for even a few upgrades on the antique elements of the design, then it'd be worth it.
Slashdot Mirror
And, no, the patent is not valid. Such devices have existed for thirty years, albeit mostly in research labs and Universities. Prior art is prior art, however, and Sony needs only one example that is close enough to invalidate the patent.
It is just possible that this lawsuit is funded by Microsoft or some other console manufacturer, in an effort to throw legalistic mud at Sony's console and thus boost sales of their own. Same way SCO tried to boost their sales by throwing legalistic mud at Linux. I would not rule that out just yet, as this may well impact sales. People aren't going to want to buy a console they can't do anything with and will have to destroy a few weeks later under court orders.
If a desktop-ish OS were to be used, LynxOS (a Linux offshoot that has some respectable avionics certification) would not be a bad choice. I'd be a little concerned about vanilla Linux, but if you stuck to the better-tested components and ripped the rest out, then really worked on validating what was left to a high standard, I'd consider it a doable option.
Haiku might also be workable, with some work, as it's designed to support a very large number of independent activities - exactly the kind of environment you'd face in a car.
Can you give me something at least a little challenging? A canal that runs uphill, a viaduct that can span a couple of cities in an earthquake zone with a 100+ year warranty, a Michelangelo-style work on the inside of a glacier - you know, something hard. Something that might be a bit more challenging than merely keeping something safe from the elements. Unfortunately, those were all done in that general area, so you'll need to pick something else.
Last I heard, there were only three being exhibited, but that doesn't mean there aren't more of them. And "left to rot" is correct - they've all suffered severe damage from corrosion, weathering, assorted plants and animals, etc. The one on loan from the Smithsonian was almost beyond repair, according to some reports, with massive amounts of replacement material being required. You treat anything on loan from me like that and you'd be making a close-up inspection of Mount St. Helens' current mountain-building activity. But because it's merely a stupendously rare internationally significant historic artifact marking the entry into genuine manned space travel, who gives a damn if it's destroyed by insects, squirrels and pollution? Apparently not the center it is parked in.
The ONLY workable system is to keep all individual votes separate in both forms and re-tally them independently. It makes no difference then whether the paper copy is identical to the corresponding electronic copy, if the grand totals for each add up to the same value. In which case, why would it matter if the sequences were not identical, provided the totals were?
However, it would be easy enough to have even that possibility checked for. Have each vote marked with a 4096-bit string that is either totally random or some sort of extremely strong cryptographic hash of the vote plus time plus other data. All you need is values that are unique and preserve anonymity. (They can't therefore be sequential or trivially derived from things that are knowable.) This would be included in both the electronic copy and the paper copy. If the extra checks are needed, simply look for unique vote IDs that exist in one system but not in the other, or for unique vote IDs where the two versions of the vote are not the same.
Is there anything else that can be done? Sure. Public key encryption can be used to digitally sign things, and the public and private keys should be kept apart. Simply put on each voter registration card an electronic tag with a public key. Two copies of the private key are made. One is at the central voting office and one is kept with independent electoral observers. When a vote is cast, the vote + unique vote ID is digitally signed. When a discrepancy is found between what is electronically cast and what is on paper, you check the votes with the suspect unique vote IDs. For each of those votes, exactly one private key (owner unknown) should match, and vice versa. Just run through the keys until one matches, then mark that key as used. If a key needs to be used more than once, then you have evidence of attempted ballot box stuffing and you can remove all votes signed with the same key.
Ok, anything else? Sure. When a vote is electronically cast, one copy goes to the central voting office (votes should NEVER be counted at the local voting center) and one copy should go to the independent observers. Both copies should be encrypted using public key encryption per voting center - not only to allow counting by center but also to validate that no center is getting more votes than voters and to prevent man-in-the-middle attacks. The independent observers should have statutory powers to challenge an election result if they detect inconsistencies, errors in procedure, flawed tallies and failures within the official body to deal with errors.
However, we now hit the hardest problem of all. As Plato noted in his "Republic", voter stupidity is the greatest flaw in democracy. When you look at the amount spent on education - just under one point three billion - it seems like a lot. When you examine how much that works out to, it's about $50 per person per year. Dunno about you, but $50 isn't going to do much educating. I doubt it'll even cover getting people familiar with a new voting system. If the only way democracy can work is through an enlightened populace, then no voting system will cure the flaws present in America.
Sure, but then I have always been a little.... odd.... The doctors say they can treat most of it, but my insurance won't cover the surgery needed to stop turning into a bat.
Written proof would be hard to come by, but yes I have received verbal confirmation from one of the developers that the project is continuing.
I disagree. A zombie army will come in handy when fighting patent battles. Lawyers can't suck the lifeblood from a zombie, because it doesn't have any.
A full reassembly by sniffing would also need to drop retransmitted packets and support all common encapsulation techniques. You're also talking about a LOT of storage and absolutely no way to sensibly organize the volume of data collected. That's the problem with data saturation - there are no database or data processing techniques capable of handling it. I was talking to one of the top Ingres software/network gurus at OSCON yesterday - apparently even just the total information awareness project is staggering under the sheer weight of information that no system yet designed can handle. If the data is unsearchable, unsortable and unprocessable, then to all practical intents and purposes, it doesn't exist.
(Given that last link, expect the RIAA to become part of Homeland Security.)
The Japanese have gigabit with IPv6 to the home already, but this makes that look like dial-up in comparison.
Given the way people wire this office? I'd want a network that supports at LEAST 40 kilometers, just to reach the server in the next room!
Dell doesn't have fans on Slashdot, but I doubt they're doing that badly in Slashdot League Division 1, and they seem to be improving on the Good Idea/Bad Idea ratio. Of course, that could just be a phase they're going through too.
Sun Tzu's classic document "Art of War" makes it very clear that you should NEVER have a politician actually commanding the armed services. The same logic goes for all other departments. Politicians are very good for looking at the big picture (well, in theory) in a way that specialists in individual fields cannot. That makes them good for determining priorities, allocating resources, setting long-term objectives, etc. But once they have issued those decisions, the rest should be entirely left to those who are competent in the field -- with one exception. Governing entitles politicians to penalize those who violate the rules necessary for a coherent organization.
The modern idea that politicians should be in control is a bastardization of the entire concept of a democracy or republic. Plato's Republic is a little dated, but does explain the difference between a ruling class and a governing class. This is an important distinction and one that many have apparently forgotten. Rulers rule. They impose. That is their nature, that is their job. If that is how you see American politics, then you are saying America has an elected monarch. (I believe the archaic term is Bretwalda, and yes elected kings have existed throughout history.) Governors govern. If the populace is the clay and the civil service are the artists, the government is nothing more than an art critic sponsoring the latest work. Nothing more.
Now, personally I don't believe that quality government exists. Here, there, or anywhere. I also generally believe that most existing Governments in the world are indeed elected monarchies... with the rest being hereditary monarchies, dictatorships and fiefdoms of various sorts. Despite the roots of constitutional law being over 5,000 years old, the notions of democracy reaching back over 2,500 years and the concept of politics as a science being studied and researched for many centuries, I can recall no time in history or in the modern world where anyone has actually applied any of these ideas.
To me, the question boils down to this. If everyone in America treats the Federal Government like a kingdom and the States like princedoms (yes, the term does exist), why not cut to the chase and cut costs at the same time by declaring it such? If people truly, honestly, believe that's what they have, then what are they going to miss by making it official? If, however, you believe that the Government is truly restricted to governing and nothing else, then you not only should imagine the government spending tax money without controlling how it is spent, you should require it.
That assumes sequential coherence. Due to latencies involved, you're more likely to get korf.
Slashdot has thrashed X many a time, so I'll focus on the kernel's hangups. To the best of my knowledge, the PPS patches bit-rotted and nobody wrote a new implementation. This limits the timer resolution Linux can handle. The kernel burns cycles on polling. The kernel context switch and userspace-to-kernelspace copying are damn expensive, which matters a lot if you're making lots of calls to the sound and graphics layers. Anything that could reduce that would help. (Intelligent graphics cards are obviously faster, but most desktops don't have GPUs, let alone GPUs of the kind of power you'd want.)
The current President's practices are, admittedly, much more grievous than many of the Thatcher-era or Blair-era quangos - generally they didn't create such groups with the power to seek (or bestow) death sentences. Well, apart from a few dodgy shoot-to-kill types, who were investigated. Well, would have been, but Stalker got kicked off the job for political reasons. However, most of the sordid scandals have involved relatively petty financial irregularities. Both Thatcher and Blair sold honors, for example, and both "fixed" illegal arms contracts.
The problem being declared is fictional, the problem being solved (the stupidity of marketroids and execs) is not. And, yeah, I've no objection to spending a few minutes on a shell script if it means a sales guy is taken to the vets and fixed.
The appropriate professional experience for a President - or any politician - is as a liar, con-artist or conjurer. I can't think of any political types - in America or Europe - who failed to meet these requirements other than the former BBC war correspondent Martin Bell (who later revealed that British parliament was far more hazardous to human health than any war zone).
I followed you up to that "simple" part. The rest of the sentence, however, is giving me problems.
As far as the reporting mechanism is concerned, it is incapable of distinguishing between slashdot.org on the Internet and an identically-named, identically-numbered machine on a private network. As the fictional users are randomly changing, along with any registration info, from a pool that was collected earlier, there would be no possible way for the stat collecting firm to identify which users are real (on the real Slashdot) and which are fake (on the fake Slashdot).
What's more, because these are done on spare boxes, there is no overhead placed on the real site and the real servers are not burdened with pointless traffic.
Once you get an idea for how much fake isolated hits you can get, you then sell a hit-booster kit to other companies. They get a similar setup (fake server with real name, and a stress-testing platform with fake Alexa users) at so many dollars per fake increment on their stats. It should be a good money-spinner, if the unit increments are large enough.
Telnet? Telnet?!? You cheeky young sods. In my day, we used Kermit! Both ways! And we were glad for it!
Spork.
Apparently you didn't notice estimates for 2006, which the National Autistic Society verifies as being 1:100 - somewhat higher than your 36:10000. In fact, how the hell did you get that number from the NAS, when the NAS' own figures (Estimated prevalence rate in the UK, given in the paper linked to above) are stated as being current and a replacement for the 1997 figures you cite.
However, you also managed to miss this little gem from the Autism Research Center at Cambridge University.
Sorry if I sound a little annoyed, but if you're going to question my figures, you need to do better than decade-old discredited reports. I don't mind people challenging my views - I enjoy the intellectual challenge - but I do object to put-downs that are out-of-date, based on flawed methods and demonstrate a lack of site-searching skills. Please! Even in D&D, I'm at a high enough level that the DM has to give the monsters PhDs in physics and advanced library science. (Ever been attacked by a goblin wielding a +20 Shakespere of Doom?)
(For the record, I was diagnosed in the US, by two independent doctors.)
These days, diagnosis by a doctor based on external symptoms is wholly unnecessary - in theory. fMRI scans will reveal Asperger's and all other autistic spectrum disorders very nicely with a very high level of accuracy. The problem with such tests is not accuracy but expense and difficulty in finding a place equipped to carry them out. Neurologists who are sufficiently far up on such diagnostic methods, given that they've only circulated from conferences in the past couple of years, are extremely rare. So whilst a symptom-based diagnosis is not required, a mechanistic diagnosis is unlikely to be in anyone's near-term future.
(However, if Slashdotters do encounter a research group wanting to verify the published results, I'd say go for it. This would give you as close to definitive proof as you can ever hope for in medical science.)
Asperger's is problematic in that there is no cure and no therapy for most of it. It is also genetically as harmful as the gene that produces Manx cats. The gene for Manx cats is additive, so eventually not only does the tail go, but so does a large chunk of the spinal cord. The gene for the Autistic Spectrum seems to be similar - if your parents both are on the Autistic Spectrum, you will be further down the spectrum. Eventually, you have nothing but mentally deformed cabbages. Asperger's can be beneficial and should be utilized as such. Play to your strengths, not your weaknesses. However, it is also a very dangerous genetic trait and should not be trusted too much. Like a Ring of Power, over-use will lead to betrayal.
Nobody even needs to actually mass-produce the buggers - Transmeta's first batch didn't even work right and had to be recalled - they just need to scare Intel into thinking they could. So long as Intel believes they are the best, they will do nothing. The second they think they're not (even if it's just a skillful illusion) is the second they'll modernize the flawed part of their design.
Oh, eventually they'll catch on, as these mysterious competitors never actually appear, but if it's good for even a few upgrades on the antique elements of the design, then it'd be worth it.