Does anyone besides me read this as "F-ing Con"? Maybe my problem is that seems like a good description of Transmeta's current business model.
Re:Chimerism doesn't create false positves
on
Chimera Twins Story
·
· Score: 1
There was, in fact, a case in England some years ago which hung on the fact that the rapist was a blood chimera: his blood test came back negative for a match. It was only when the investigators got a sperm sample that they could find a DNA match.
Did anybody notice that the original submittor managed to get this onto the front page by ignoring the fact that chimerism would decrease the likelihood of a match, rather than increasing it? -1, troll submission.
If you're going to quote text, you ought to quote the relevant text: the definition of a person to whom the Third Geneva Convention applies.
Here's the main definition:
A. Prisoners of war, in the sense of the present Convention, are persons belonging to one of the following categories, who have fallen into the power of the enemy:
1. Members of the armed forces of a Party to the conflict as well as members of militias or volunteer corps forming part of such armed forces.
2. Members of other militias and members of other volunteer corps, including those of organized resistance movements, belonging to a Party to the conflict and operating in or outside their own territory, even if this territory is occupied, provided that such militias or volunteer corps, including such organized resistance movements, fulfil the following conditions:
(a) That of being commanded by a person responsible for his subordinates;
(b) That of having a fixed distinctive sign recognizable at a distance;
(c) That of carrying arms openly;
(d) That of conducting their operations in accordance with the laws and customs of war.
The section goes on, but the subsequent passages either speak to edge cases.
Key is, a covert enemy agent meets none of the four tests for being a prisoner of war. In that case, there's no question about whether or not that agent is covered by the provisions of the Convention; he or she is not. Irregular combatants may or may not be, but generally would not be covered. The foreign combatants in Afghanistan directly associated with Al Quaeda were clearly not covered: they were not commanded by a responsible officer, they wore no distinctive signs, they concealed their weapons, and they did not conform to the standard laws and customs of warfare (including the Third Geneva Convention, which forbids the taking of hostages and direct attacks on civilians, both of which many of the GB detainees had done.)
In short, GB may be wrong, and is a PR disaster, but it is not illegal, no matter what HRW wants you to believe.
Recall that Usama (or Osama) bin (or ibn) Laden had been involved in a number of attacks against US interests over the years. I want you to think back to August 7, 1998. On that morning, truck bombs were detonated outside the US Embassies in Kenya and Tanzania. 22 Americans died, and 223 Kenyans and Tanzanians died. Over the next few years, the second tier of people who had prganized the bombings were captured and extradited to New York to face trial.
They were tried as members of a terrorist organization called "Al Quaeda", which, under the direct and indirect supervision of bin Laden, planned and executed the 1998 bombings. So what? Well, they were convicted in August of 2001, and the transcripts of their trials are public documents. You can find a collection of them here.
Whether or not he was associated with 9/11, the taliban were sheltering him from justice, and they knew it.
Actually, you are right. You wouldn't have been a terrorist. You'd have been a covert agent of an enemy power -- a spy. You would not have been held as a POW; you would have been held as an irregular combatant. Guess what? The Geneva conventions don't protect irregular combatants. Combatant nations are not legally bound to return irregular combatants to their countries of origin when conflict ends.
If you weren't shot out of hand, you'd have spent the rest of your life in a French jail, along with the other collaborators.
No. What you're told is that you can't coerce my children to participate in the practice of your religion -- or, rather, your sect's abuse of religion. Of course, I suppose that I could ask the new Bishop from my branch of the Church to come speak on the tradition of biblical misinterpretation regarding homosexuality to your kids school, if you really want people's religion in the public sphere.
My freedom from your religion is the same right as your protection from mine. If you won't accept the first, then you don't get the second.
I've heard that so many times referring to Microsoft as "they". Now Linux is "they". The open source community is now to the "fighting" stage against SCO. That would mean that the next stage is SCO winning.
Well, pch is certainly an issue, but it's the quality of the resulting object code which is a problem. mingw is a good example: its support for generic 486-and-later x86 assembler is pretty limited, and we usually don't have the option of compiling for recent architectures only. The resulting object code is quite bloated. That's fine for a small benchmark, but for a larger body of code, it's a disaster. The smaller size of the msvc/icc binaries has significant perf consequences.
In some sense, the problem is that you're asking the wrong question. There are significant limitations in the C/C++ language definition compliance in MSVC6 (although as of MSVC7, we're way ahead of everyone else). We can work around those: a Turing machine's a Turing machine, no matter how you represent the Godel numbers of the programs you run. As long as I can write "for (int i = 0; i SEVENTEEN; ++i)" I can make do, even if the scope of the variable i is wrong. I'll just learn to talk the MSVC dialect when I'm talking to VC. What comes out the back is much more important, and there, the VC code generator is only matched by the Intel code generator plug in.
I'm an employee. There's no reason for me to respond anonymously: the corporate policy is well publicized.
Internally, corporate policy has always been that we can use whatever tools we want, provided that they serve the purposes of our jobs best. The only restriction under which we work is that developers or other people who have direct check-in rights to any of our trees (think "committer privileges"; it's the closest thing in the FOSS world) are not permitted to examine code released under GPL or any other viral license.
So, yes, for testing interop, we have a lot of Linux/Apache boxes around. We have a lot of Perl. (We've been supporting ActiveState for years, after all.) If there were a competitive FOSS compiler available, I'm sure that some groups would use it. There isn't. We've certainly had teams do comparitive analyses.
One of the developers in my group is a forty-something year old guru who run XEmacs on his main dev machine. Whenever he reformats his machine, he does a pure binary install and deletes the.el files to keep things clean. His attitude is that it's easier for him to keep using a tool which he recommends other developers avoid than it is to learn a newer and more efficient tool. I've asked him if he ever received any pressure to change. His answer was, "Never. [Our boss] doesn't care what we use to write code; he cares that we write code."
Maybe he's drunk too much of the KoolAid -- but my experience tracks his. Think about it. Why would we care? If one of our gurus is more productive using XEmacs, that is at worst a data point for the Visual Studio folks.
When I read this, I'm reminded of the SPHINX project at CMU in the mid 80's. Kai-Fu Lee was a doctoral student at CMU in computer science. His advisor set him to evaluating the performance of the (clearly inferior) statistical SR systems that IBM was touting. It was a throw-away project; his advisor just wanted some numbers to compare his rule-based system against. The linguists had clearly shown that the irregularities of human speech required deep knowledge of the phonology, syntax, and sematics of the language being spoken, but the projectg leader needed a benchmark to measure against.
Lee's toy project, SPHINX, won the DARPA competition that year. The highest scoring rule-based system came in fifth. What the linguists "knew" was wrong.
The example you gave is another example of the linguists not know as much about statistics as they think. The corpora used for statistical translation include examples of idiomatic usages. Idiomatic usage is highly stereotypical, so the Viterbi path through an N-gram analysis captures such highly linked phrases with high accuracy.
Consider this case: someone purports to sell you a watch. It later turns out that the watch was stolen, and it is impounded by the police, and returned to its "real" owner. If you had bought that watch, then, yes, you'd own it, and the cops couldn't do that. Problem is, you see, that you did not "buy" the watch. The person who took money from you in exchange for the possession of the watch neither owned it nor had a contract to transfer ownership. He or she could not, therefore, sell it to you.
In the case of the counterfeit HP book, the same argument applies. The bookseller did not have the authority to sell you the book, and so you don't own it. It's arguable that if you own the atoms which comprise the book (although that would be harder to argue than you think), but you certainly don't own that arrangement of those atoms, and you don't have a valid license to make a copy of the information that arrangement encodes.
The impress of the (allegedly) illegal copy of Linux has exactly the same properties. You own the disk, but you don't own that arrangement of atoms on the disk. You don't have the right to have it.
Not quite. Look carefully at the first part of Section 107.
Notwithstanding the provisions of section 106, it is not an infringement for the
owner of a copy of a computer program
The key word is "owner". You are not the owner of an illegal copy of a computer program. You are a person who is in the possession of misappropriated goods.
Moglin is full of sh...I mean, he's being disingenous. How did you get the binary onto the computer in the first place? You copied it onto that machine's storage, of course. That copy, right there, is infringing.
The DMCA contains an expemption for installation of software. That exemption is carefully tailored, though: it only applies to software which you have obtained legitimately. If the person from whom you got the original installation media didn't have a license to distribute them, then that exemption doesn't apply. It's a situation of "buyer beware".
Hold it right there, sir. Let's look back here at your point...was it 4?...yes, four. This notion of "consideration". Now, I may just be a southern boy, but I seem to remember that meant "valuable consideration" right?
So would you be good enough to tell the court how much I would need to pay to get a copy of this here "Leenicks" thing? Seems to me that you just said I could get a copy for nothing, right? That, in fact, Torvalds sends out copies for free, too?
</drawling southern lawyer voice>
Aren't you effectively saying that means there's no value to the "right to distribute the code" -- if I need a copy, I can get it directly from the manufacturer gratis. A contract requires the exchange of valuable consideration -- and there's none of that.
Your honor, there is no contract involved here. The defense moves for immediate dismissal of this action, with prejudice.
[I]t's clear copyright infringement to make an
unlicensed copy of a copyrighted work on your hard disk.
(Emphasis added.) Your chum held the copyright on the file, and hence had a clear license to upload and/or download the file in question. If I had done the same thing, it would have been an instance of infringement, since I don't have a license to d/l that file. The copy I made on my disk by the act of downloading (whether or not I kept it) would have been infringing.
No. Downloading from Napster/etc. is patently illegal -- it's clear copyright infringement to make an unlicensed copy of a copyrighted work on your hard disk. The DCMA includes a special exemption to that: if you have software which requires a copy be made in order to run the software, then you may make that copy. That way, you don't need a license to install software you've acquired through legal means.
Basically, if there's code in the Linux kernel which isn't properly licensed, then anyone running a copy of the kernel is infringing.
Did you even read the comment you were responding to? "Proper database normalization"? "Standardized software platforms"? "Open specifications"?
It is hard to imagine a better example of an open specification than a government project. The documentation is open, and the specifications are designed by the customer to meet its requirements. More than that, the customer in this case has a LOT of clout -- governments aren't likely to be going anywhere anytime soon.
That's fine -- but the cost of reengineering a solution are huge, and the risks tremendous. And what about managing the transition? And what if the specs need to change? Which is cheaper and provides better customer service: hacking on the database in a way that wasn't ever intended, or replacing the database?
Fine, you say, we'll add some extra fields for future expansion. What are you going to say when somebody asks, suspiciously, "What do you mean 'future expansion'? What are you planning on hiding there?" Remember that once the database is filled, members of the public won't get to see its contents, as they become confidential information. Every loony with a tin-foil hat will decide that he or she knows exactly what kind of "future expansion" you're planning, and they'll want to know where you're hiding the budget for the black helicopters. The rational minded libertarians will talk about how you're planning to use that data to exploit taxpayers. And so on.
And some damn fool open sourcerer will spout off about "how you could do better with MySQL" -- even though he knows no more about the constraints on your database than the loonies do. (What consequences does HIPPA have for the selection of database platforms in the DHHS? Do the special requirements placed on VA systems mean that the sequrity levels accorded to that data need to be raised? How important is interoperation with other government departments, such as Social Security, and is that level of interop acheivable if we do any hand coding? If not, what commercial solution do we go with which minimizes the risk that things will fall apart?)
Please put your tin-foil hat away. The incorrect use of humor will not flag anyone for review as a potential terrorist. There is no reason to be concerned that we will interfere with any humor-related deviance. It is only in those cases where individuals with perverted senses of so-called humor that pose a threat to our national security (as determined by our objective and reproducible criteria), and who aver themselves unwilling to participate in our voluntary humor-retraining camps, who will be marked for review. In order to reduce the number of individuals whose privacy will be sacrificed to review, we will use only publicly available data. In order to incentivize those who will be encouraged to attend humor-improvement camps, we intend to locate them in tropical locations near to the ocean, but not on US territory.
Frodo falls under the spell of the Ring, and rises to become the new Dark Lord. With the help of Gandlaf, Frodo leads a successful invasion of the Undying Lands. Darkness rules eternally.
Belive it or not, Pocket PC prices have been falling steadily over the last few years. The JVC device costs less that $500 and includes 802.11b and several card slots, but a first generation iPaq cost upwards of $550 before you bought the expansion sled. To spend that kind of money on a Pocket PC now, you'd need to buy a Phone Edition version, and even that's getting pretty hard.
This is because of the economics of the device market more than anything else. Customers will pay a premium for useful features like 802.11b, but the BOM cost is heavily driven by the screen, the battery, and the CPU. It's more profitable to take a sloppy but reliable reference design and simply slap a few premium features on it than it is to do the extensive optimization of circuit board layout and power supply to make a profit at the low end.
If you're going to use ground temperature, though, why not go the whole way, and use a geothermal heat pump? Run non-reactive tubing through your yard (or vertically, if you want to), run a liquid coolant through that tubing, and use the 17 C ambient earth temperature to heat and cool your PC without drilling holes in your foundation.
And while you're at it, you can also keep your dwelling at a temperature not too far from the temperature at which your computer is kept. And, as a bonus, it'll cost you about $0.50/day to maintain this constant temperature.
Mmmm. Saving two thousand dollars a year of heating and cooling costs -- better add a few more meters of tubing there, to handle the new computers you'd be able to afford.
Look at the traditional FSF claim that software using Bison as a parser is GPL. I'm merely parroting the standard argument offered there and reading its full implications.
Realistically, I don't care if RMS and Eban Moglin don't "mean" that with GPL. An agreement means what a malevolent third party wants it to mean, not what a benevolent author wants it to mean. Just because FSF says "Oh, we'd never read the agreement to mean that" doesn't mean that a third party lawyer wouldn't. When you're interpreting a contract, don't ask yourself what Richard Stallman would read into it -- ask yourself what Darl McBride would read into it.
Slashdot Mirror
Of course, it doesn't actually support all, or even most, Excel functionality.
It lacks Visual Basic support. And OLE support.
Does anyone besides me read this as "F-ing Con"? Maybe my problem is that seems like a good description of Transmeta's current business model.
There was, in fact, a case in England some years ago which hung on the fact that the rapist was a blood chimera: his blood test came back negative for a match. It was only when the investigators got a sperm sample that they could find a DNA match.
Did anybody notice that the original submittor managed to get this onto the front page by ignoring the fact that chimerism would decrease the likelihood of a match, rather than increasing it? -1, troll submission.
Here's the main definition:The section goes on, but the subsequent passages either speak to edge cases.
Key is, a covert enemy agent meets none of the four tests for being a prisoner of war. In that case, there's no question about whether or not that agent is covered by the provisions of the Convention; he or she is not. Irregular combatants may or may not be, but generally would not be covered. The foreign combatants in Afghanistan directly associated with Al Quaeda were clearly not covered: they were not commanded by a responsible officer, they wore no distinctive signs, they concealed their weapons, and they did not conform to the standard laws and customs of warfare (including the Third Geneva Convention, which forbids the taking of hostages and direct attacks on civilians, both of which many of the GB detainees had done.)
In short, GB may be wrong, and is a PR disaster, but it is not illegal, no matter what HRW wants you to believe.
Recall that Usama (or Osama) bin (or ibn) Laden had been involved in a number of attacks against US interests over the years. I want you to think back to August 7, 1998. On that morning, truck bombs were detonated outside the US Embassies in Kenya and Tanzania. 22 Americans died, and 223 Kenyans and Tanzanians died. Over the next few years, the second tier of people who had prganized the bombings were captured and extradited to New York to face trial.
They were tried as members of a terrorist organization called "Al Quaeda", which, under the direct and indirect supervision of bin Laden, planned and executed the 1998 bombings. So what? Well, they were convicted in August of 2001, and the transcripts of their trials are public documents. You can find a collection of them here.
Whether or not he was associated with 9/11, the taliban were sheltering him from justice, and they knew it.
Actually, you are right. You wouldn't have been a terrorist. You'd have been a covert agent of an enemy power -- a spy. You would not have been held as a POW; you would have been held as an irregular combatant. Guess what? The Geneva conventions don't protect irregular combatants. Combatant nations are not legally bound to return irregular combatants to their countries of origin when conflict ends.
If you weren't shot out of hand, you'd have spent the rest of your life in a French jail, along with the other collaborators.
No. What you're told is that you can't coerce my children to participate in the practice of your religion -- or, rather, your sect's abuse of religion. Of course, I suppose that I could ask the new Bishop from my branch of the Church to come speak on the tradition of biblical misinterpretation regarding homosexuality to your kids school, if you really want people's religion in the public sphere.
My freedom from your religion is the same right as your protection from mine. If you won't accept the first, then you don't get the second.
Then they laugh at you
Then they fight you
Then you win. -- Ghandi
I've heard that so many times referring to Microsoft as "they". Now Linux is "they". The open source community is now to the "fighting" stage against SCO. That would mean that the next stage is SCO winning.
Well, pch is certainly an issue, but it's the quality of the resulting object code which is a problem. mingw is a good example: its support for generic 486-and-later x86 assembler is pretty limited, and we usually don't have the option of compiling for recent architectures only. The resulting object code is quite bloated. That's fine for a small benchmark, but for a larger body of code, it's a disaster. The smaller size of the msvc/icc binaries has significant perf consequences.
In some sense, the problem is that you're asking the wrong question. There are significant limitations in the C/C++ language definition compliance in MSVC6 (although as of MSVC7, we're way ahead of everyone else). We can work around those: a Turing machine's a Turing machine, no matter how you represent the Godel numbers of the programs you run. As long as I can write "for (int i = 0; i SEVENTEEN; ++i)" I can make do, even if the scope of the variable i is wrong. I'll just learn to talk the MSVC dialect when I'm talking to VC. What comes out the back is much more important, and there, the VC code generator is only matched by the Intel code generator plug in.
I'm an employee. There's no reason for me to respond anonymously: the corporate policy is well publicized.
.el files to keep things clean. His attitude is that it's easier for him to keep using a tool which he recommends other developers avoid than it is to learn a newer and more efficient tool. I've asked him if he ever received any pressure to change. His answer was, "Never. [Our boss] doesn't care what we use to write code; he cares that we write code."
Internally, corporate policy has always been that we can use whatever tools we want, provided that they serve the purposes of our jobs best. The only restriction under which we work is that developers or other people who have direct check-in rights to any of our trees (think "committer privileges"; it's the closest thing in the FOSS world) are not permitted to examine code released under GPL or any other viral license.
So, yes, for testing interop, we have a lot of Linux/Apache boxes around. We have a lot of Perl. (We've been supporting ActiveState for years, after all.) If there were a competitive FOSS compiler available, I'm sure that some groups would use it. There isn't. We've certainly had teams do comparitive analyses.
One of the developers in my group is a forty-something year old guru who run XEmacs on his main dev machine. Whenever he reformats his machine, he does a pure binary install and deletes the
Maybe he's drunk too much of the KoolAid -- but my experience tracks his. Think about it. Why would we care? If one of our gurus is more productive using XEmacs, that is at worst a data point for the Visual Studio folks.
When I read this, I'm reminded of the SPHINX project at CMU in the mid 80's. Kai-Fu Lee was a doctoral student at CMU in computer science. His advisor set him to evaluating the performance of the (clearly inferior) statistical SR systems that IBM was touting. It was a throw-away project; his advisor just wanted some numbers to compare his rule-based system against. The linguists had clearly shown that the irregularities of human speech required deep knowledge of the phonology, syntax, and sematics of the language being spoken, but the projectg leader needed a benchmark to measure against.
Lee's toy project, SPHINX, won the DARPA competition that year. The highest scoring rule-based system came in fifth. What the linguists "knew" was wrong.
The example you gave is another example of the linguists not know as much about statistics as they think. The corpora used for statistical translation include examples of idiomatic usages. Idiomatic usage is highly stereotypical, so the Viterbi path through an N-gram analysis captures such highly linked phrases with high accuracy.
Consider this case: someone purports to sell you a watch. It later turns out that the watch was stolen, and it is impounded by the police, and returned to its "real" owner. If you had bought that watch, then, yes, you'd own it, and the cops couldn't do that. Problem is, you see, that you did not "buy" the watch. The person who took money from you in exchange for the possession of the watch neither owned it nor had a contract to transfer ownership. He or she could not, therefore, sell it to you.
In the case of the counterfeit HP book, the same argument applies. The bookseller did not have the authority to sell you the book, and so you don't own it. It's arguable that if you own the atoms which comprise the book (although that would be harder to argue than you think), but you certainly don't own that arrangement of those atoms, and you don't have a valid license to make a copy of the information that arrangement encodes.
The impress of the (allegedly) illegal copy of Linux has exactly the same properties. You own the disk, but you don't own that arrangement of atoms on the disk. You don't have the right to have it.
Moglin is full of sh...I mean, he's being disingenous. How did you get the binary onto the computer in the first place? You copied it onto that machine's storage, of course. That copy, right there, is infringing.
The DMCA contains an expemption for installation of software. That exemption is carefully tailored, though: it only applies to software which you have obtained legitimately. If the person from whom you got the original installation media didn't have a license to distribute them, then that exemption doesn't apply. It's a situation of "buyer beware".
See? It's true! Linux does eqaul Kommunism!
Doesn't "???? = #2" follow from "#2 = ????"?
Hold it right there, sir. Let's look back here at your point...was it 4?...yes, four. This notion of "consideration". Now, I may just be a southern boy, but I seem to remember that meant "valuable consideration" right?
So would you be good enough to tell the court how much I would need to pay to get a copy of this here "Leenicks" thing? Seems to me that you just said I could get a copy for nothing, right? That, in fact, Torvalds sends out copies for free, too?
</drawling southern lawyer voice>
Aren't you effectively saying that means there's no value to the "right to distribute the code" -- if I need a copy, I can get it directly from the manufacturer gratis. A contract requires the exchange of valuable consideration -- and there's none of that.
Your honor, there is no contract involved here. The defense moves for immediate dismissal of this action, with prejudice.
(Emphasis added.) Your chum held the copyright on the file, and hence had a clear license to upload and/or download the file in question. If I had done the same thing, it would have been an instance of infringement, since I don't have a license to d/l that file. The copy I made on my disk by the act of downloading (whether or not I kept it) would have been infringing.
No. Downloading from Napster/etc. is patently illegal -- it's clear copyright infringement to make an unlicensed copy of a copyrighted work on your hard disk. The DCMA includes a special exemption to that: if you have software which requires a copy be made in order to run the software, then you may make that copy. That way, you don't need a license to install software you've acquired through legal means.
Basically, if there's code in the Linux kernel which isn't properly licensed, then anyone running a copy of the kernel is infringing.
Did you even read the comment you were responding to? "Proper database normalization"? "Standardized software platforms"? "Open specifications"?
It is hard to imagine a better example of an open specification than a government project. The documentation is open, and the specifications are designed by the customer to meet its requirements. More than that, the customer in this case has a LOT of clout -- governments aren't likely to be going anywhere anytime soon.
That's fine -- but the cost of reengineering a solution are huge, and the risks tremendous. And what about managing the transition? And what if the specs need to change? Which is cheaper and provides better customer service: hacking on the database in a way that wasn't ever intended, or replacing the database?
Fine, you say, we'll add some extra fields for future expansion. What are you going to say when somebody asks, suspiciously, "What do you mean 'future expansion'? What are you planning on hiding there?" Remember that once the database is filled, members of the public won't get to see its contents, as they become confidential information. Every loony with a tin-foil hat will decide that he or she knows exactly what kind of "future expansion" you're planning, and they'll want to know where you're hiding the budget for the black helicopters. The rational minded libertarians will talk about how you're planning to use that data to exploit taxpayers. And so on.
And some damn fool open sourcerer will spout off about "how you could do better with MySQL" -- even though he knows no more about the constraints on your database than the loonies do. (What consequences does HIPPA have for the selection of database platforms in the DHHS? Do the special requirements placed on VA systems mean that the sequrity levels accorded to that data need to be raised? How important is interoperation with other government departments, such as Social Security, and is that level of interop acheivable if we do any hand coding? If not, what commercial solution do we go with which minimizes the risk that things will fall apart?)
Please put your tin-foil hat away. The incorrect use of humor will not flag anyone for review as a potential terrorist. There is no reason to be concerned that we will interfere with any humor-related deviance. It is only in those cases where individuals with perverted senses of so-called humor that pose a threat to our national security (as determined by our objective and reproducible criteria), and who aver themselves unwilling to participate in our voluntary humor-retraining camps, who will be marked for review. In order to reduce the number of individuals whose privacy will be sacrificed to review, we will use only publicly available data. In order to incentivize those who will be encouraged to attend humor-improvement camps, we intend to locate them in tropical locations near to the ocean, but not on US territory.
Frodo falls under the spell of the Ring, and rises to become the new Dark Lord. With the help of Gandlaf, Frodo leads a successful invasion of the Undying Lands. Darkness rules eternally.
Belive it or not, Pocket PC prices have been falling steadily over the last few years. The JVC device costs less that $500 and includes 802.11b and several card slots, but a first generation iPaq cost upwards of $550 before you bought the expansion sled. To spend that kind of money on a Pocket PC now, you'd need to buy a Phone Edition version, and even that's getting pretty hard.
This is because of the economics of the device market more than anything else. Customers will pay a premium for useful features like 802.11b, but the BOM cost is heavily driven by the screen, the battery, and the CPU. It's more profitable to take a sloppy but reliable reference design and simply slap a few premium features on it than it is to do the extensive optimization of circuit board layout and power supply to make a profit at the low end.
If you're going to use ground temperature, though, why not go the whole way, and use a geothermal heat pump? Run non-reactive tubing through your yard (or vertically, if you want to), run a liquid coolant through that tubing, and use the 17 C ambient earth temperature to heat and cool your PC without drilling holes in your foundation.
And while you're at it, you can also keep your dwelling at a temperature not too far from the temperature at which your computer is kept. And, as a bonus, it'll cost you about $0.50/day to maintain this constant temperature.
Mmmm. Saving two thousand dollars a year of heating and cooling costs -- better add a few more meters of tubing there, to handle the new computers you'd be able to afford.
Look at the traditional FSF claim that software using Bison as a parser is GPL. I'm merely parroting the standard argument offered there and reading its full implications.
Realistically, I don't care if RMS and Eban Moglin don't "mean" that with GPL. An agreement means what a malevolent third party wants it to mean, not what a benevolent author wants it to mean. Just because FSF says "Oh, we'd never read the agreement to mean that" doesn't mean that a third party lawyer wouldn't. When you're interpreting a contract, don't ask yourself what Richard Stallman would read into it -- ask yourself what Darl McBride would read into it.