In a former life, I was an co-op student at Western Union (Anyone remember them? They used to actually transfer data as well as money.), and I was involved in leasing data lines from AT&T and the local telcos wherever our customers needed connections.
Some of the larger outfits were indeed interested in redundancy and paid a premium for, say, two links from NYC to SF, one by way of Chicago, and another through Houston. We frequently had trouble verifying that we were really getting independent routes.
The telcos bundled connectivity (back when a single voice line was lotsa data links [after they'd been digitized]) in a hierarchy so deep that it took days or weeks to verify that every link in the route used facilities physically separate from every other.
Why? 'Cause to Ma Bell, bandwidth is fungible. Got noise on link 37A from Manhattan to Albany? Take it out of service for maintenance, and swap in some spare bps from Manhattan to Jersey City to Albany. It's all the same....
Until a manhole floods in Jersey City, and *oops* it turns out your route to Chicago (formerly via Albany) is now cheek-by-jowl with the wire to Houston, and you're dead!
Among other things, we supposedly rented circuits with huge Do Not Reroute tags hanging all over them, but on occasion someone overlooked it, or worse, the facility two levels up the hierarchy got rerouted and our link went along for the ride unknowingly.
I wish them well---it can only be much worse with fiber optics these days.
I'm worried about the rush to fund Internet information sources from advertising---it'll end up heading the same way TV has: it's a medium to sell consumers to producers, with the content a distant second, useful only as a lure for the consumers.
Micro-payments would be an ideal way to avoid this problem, but the mechanisms to ensure reliability and anonymity, though extant, are pitifully slow in being implemented, and are further hindered by the absence of standards. I'd gladly pay, say, USD 0.0001 (a hundredth of a cent for the math-impaired) for any web page I'm served, so long as it's totally transparent to me. There are plenty of sites out there that get much less than that (read: zero). Unfortunately that's not going to happen soon.
So, how about those sites to whom I'd pay significantly more than that, maybe even an order of magnitude more, like Slashdot?
Anonymous micro-payments are good for surfing, but when I've found someplace really worthwhile, why shouldn't it offer advertising-free pages to me in exchange for cash? Are you listening, Rob?
How much do you get in advertising revenue for my presence? I bet I'd pay much more than that to rid my pages of ads. All you need to do is offer a, say USD 5.00--10.00 annual subscription (is this in the right ballpark?) to volunteers. I give you a credit-card number, you bill me annually, and when I log on, voila---no ads. If enough think this is worthwhile, you could get rid of the ads for everyone. I wouldn't mind subsidizing a few dozen ACs and free riders to get a/. which couldn't care less about the feelings of anyone other than its readers.
Of course, this raises the specter of/. dominated by the paying customers, but keep the price down, and anybody can play. Not too bad!
Thanks for a great article. Tell Jon Katz to reflect for a while on what makes this one so good:-).
...if you think in terms of decades or centuries. Do you think the x86 architecture will last forever? I've got floppies with documents and their word processor on them, but until I find a CP/M machine they won't do me much good!
(Yes, I know about simulators, but it's still not like rolling off a log.)
The E-biomed flap is highlighting a lack in U.S. (at least) society's educational/social system. We try to teach our children all manner of facts and algorithms, but I've never seen (short of college) an explicit attempt to teach the need to evaluate information, and more importantly the methods for such evaluation. As the state of the nation demonstrates, we desperately need such.
The rise of the Internet has only exacerbated a pre-existing problem, as the existance of sleazy politicians, Ponzi schemes, and homeopathic medicine amply demonstrate. Until we get up off our arses and start looking into the information we receive and the biases of those supplying it, we will be putty in the hands of scum.
But, we can't check everything ourselves, we must delegate some of the examination to others, and the question we're missing is: ``How do we evaluate the trustworthiness of an informant?''
Such an evaluation is a case of diminishing recursion: Those we already trust can point us to informants they trust, but if we're wise we invest less trust the further the chain extends. But whom do we trust de novo? I claim that the only way to develop trust is through memory of previous information that has been checked afterward.
Did ``trickle-down'' economics work? Ask yourself how your situation changed during the Reagan-Bush years, then apply what you've learned to the trustworthiness rating of those supporting it. Does/. supply unerring information? Check up on the items you're interested in and adjust your trust accordingly. Without the effort to remember and evaluate, we're no better than animated bags of prejudice.
Now, down to cases: Why do I trust the New York Times reporting more than that of the National Enquirer? Because they've been proven to be full of malarkey less often. (Too frequently, but still less often.) Why do I trust the Merck Manual more than the local herbalist's gazette? Because the Merck tries to base itself on replicable data instead of anecdote. Does that mean one's gospel and the other's fish-wrapping? No---trust is not a binary value. It runs the gamut from nearly zero (but, if a fool says the Sun rises in the East, it's no less true) to almost complete (but even I'm fallible).
We need to start teaching trust evaluation. One of the major problems in this day and age is the value placed on quick information. Evaluation and verification takes time, and information immediately available is inherently less trustworthy---something many on/. ignore. For most information, speed is not of the essence, accuracy is. The only way we can trust speedy information is if we trust the informant, and in e-info, the informant is all too often someone we've never heard of. If Alan Cox reports a kernel bug that can wipe my hard drive, I'll act on it a lot faster than if it comes from a/. AC.
But, trust also has subject limits. I'll listen respectfully to Mr. Cox on the kernel, programming, etc., but when it comes to investment information or medical treatments he's just another tyro until I've evaluated his abilities in those fields.
[This is one reason I'm happy that we have accounts available on/.---I certainly don't know who really signed up for the LTorvalds account (if such exists), but if an account holder is reasonably careful with the password, I can trust that the DonkPunch I heard from last week is the one posting today, and can apply the trust rating I previously developed for that poster to the current post.]
Thus, information evaluation is based on how long the info's been available for evaluation, and the track record of those who've evaluated it (if I'm unable to do so personally). Teach that to society, and we'll all live better (except for the scoundrels out there).
As for Dr. Varmus, it indeed sounds as if he's got his head screwed on straight. He's setting up a system where information is explicitly separated by speed of availability and depth of examination---take whichever you like, but remember to apply the appropriate assessment based on where you got it.
As people begin to pay more attention to reliability of information, we'll see more such web sites develop---they may not have the rumors that came through in the last 30 seconds, but they'll be the places you go when it really matters.
After reading the decision on the Bernstein case (see the/. article for a refresher), the typo in Stephenson's book struck me as a perfect defense.
The decision refers to the Government's claim that the functional aspects of crypto override any First-Amendment issues. (See pp. 4235, 4236, 4238n., and the dissent on pp. 4246ff.) Thus, we can export the first edition of Cryptonomicon in machine-readable form, free of obnoxious restraints, precisely because it doesn't function. A comment on the code would be nice, or a pointer to the errata, but if the code don't work, the Government can't claim it's crypto!
A reference to Nietzche's Beyond Good And Evil (translated by Walter Kaufmann) quotes it as:
Whoever fights monsters should see to it that in the process he does not become a monster. And when you look long into an abyss, the abyss also looks into you.
Would it be worth a try to let readers nominate moderators? That way, when I notice a poster whose comments are really cogent, I can mark that person as a potential moderator.
Of course, this posits that good comments come from people with the balance and judgment to moderate well. I think that's true as a first approximation.
How to use this input, however? Count up the number of times a particular poster gets nominated, and the highest numbers get moderator privileges? (`Nomination' is really voting.) Alternatively, if the count goes over some threshold, have a human look at that person's posts and decide yea or nay. (`Nomination' is just that, with other criteria for acceptance.)
I suggest giving the first mechanism a try, because there are some folks whose judgment I respect, and would like to see have some input based on their abilities rather than pure numerology. (<nose color="brown">Not that the numerology isn't a damned good shot at automating a tough call.</nose>)
nomination would allow for notably-clueful newbies to get moderatorhood
only logged-in readers can nominate (and only nominate a given poster only once)
keep an eye on who gets nominated to watch for attempts to stuff the ballot box
Maybe nominations can come only from readers in domains other than the nominee's? Perhaps only from the older 2/3 of the population? (Both of these to prevent someone from registering an bunch of IDs and performing auto-nomination.)
Originator still works for Circuit City
on
Anti-DIVX article
·
· Score: 1
That's `cause he owns it. I've read that Rick Sharp (head of Circuit City) thought up the outlines of the product, then assigned some good folks to flesh it out. I don't think he's going to fire himself.:-) (Hi, Rick!)
(Dupe so I'm logged in---posting code forgot between preview and post last time.)
That's ``...he argues against the FSF....''
on
Feature:Free Linux
·
· Score: 1
It's a perfectly good English interpretation to read your intro as ``[Christiansen] argues, with the FSF and RMS, that...'' (note the added commas), which means he agrees with them. In fact, that's what I expected to read, and was taken aback when the article said the exact opposite.:-)
English is wonderful, not least due to its openness to the well-place ambiguity, but when you're writing blurbs, make sure there are none you didn't intend.
This is a copyright suit, not a patent one. Prior art invalidates a patent, but if you have a copyright, you have rights to the expression (roughly, the way you coded it), not to the algorithm.
Have you ever been within smelling distance of a lawsuit? If not, read, for example, coverage of the Microsoft suit the DoJ's prosecuting.
I have a number of friends who are lawyers (and actually have working consciences!), and get the inside scoop from them about the operation of the law. As P. J. Plauger observed in an Embedded Systems Programming column, anytime you go to court, you've lost, regardless of the merits of your case.
If we accept licenses that allow courts to decide what the license language means (as a primary function, not just as secondary to deciding whether the license has been violated), we're signing our own death warrant. Anyone with a few bucks to rub together can bring all development to a complete halt at any time for any project using such a license.
Clue supply: Patent's aren't for developers
on
VMWare Beta Release
·
· Score: 1
To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries;....
(Emphasis added---see your local social contract for other countries.)
It's not about an inventor's ownership---there is no ownership of ideas, nor even any limits on use except as society has decided it's to its advantage to allow such.
The advantage posited is that if society offers remuneration to those who come up with new ideas, we'll get more such ideas, and civilization will advance that much more rapidly. (I'm all in favor of penicillin and contact lenses, myself.) The remuneration is in the form of a monopoly on use of the idea (for a limited time), allowing the thinker to make money on it before others get a shot at copying it.
This was a Good Thing for ideas that took years of sweat, tons of metal, and a lot of limited and expensive resources to develop, like, say, Bessemer furnaces, safety pins, and leading-shoe brakes. If the idea wouldn't be implemented without a pay-back for the thinker, society was less likely to reap its benefits. So, we say (for patents) ``Here, tell us what your idea is, document it for all to see, and in return for that openness we'll help you make money from it---not because you have any ownership of an idea, but because we'll get the benefit of having it sooner, and letting other thinkers build on it.''
The vital concept here is that patents are for the good of society, not the person with the idea. If an idea is most likely going to be developed anyway, there's no social benefit to limit use of it. Further, software is one of the most likely areas for ideas to be developed without monopoly, simply because coding is cheap, attracts thousands (millions?) of people, and is the most fun you can have with your clothes on. If one hacker doesn't generate a given idea, there'll be another one along in a minute.
Unfortunately, over the decades, people got (remained?) greedy, and specialized enough to have never studied their own history, and so began to think that an idea, of all things, could have an owner, and that it was their right to have a monopoly and control access to the idea.
Look back on the development of software: in the past 60 years is there any evidence that its progress has been hampered by people afraid to develop ideas for lack of protection? Hah! Never in the history of technology has a field advanced so rapidly or been so fecund, and most of it has been due to the open availability of ideas. Until the 1980s no one could patent software, and copyright seemed to work quite nicely, thank you. Now, unfortunately, we have people who think that because they spent a year or two on an idea, they should be allowed to prohibit anyone else from using the idea, even though it's hardly novel, or even notably hard---tedious in reducing to a useful form, but no more.
So, someone building a virtual machine (which IBM made mega-bucks on in the '60s and '70s [and may be still, for all I know]) cannot be allowed to say ``I'm the only person who can use this idea for the next 17 years'' (or is it up to 20 now?) when the same idea has occurrred to anyone who's written an emulator.
Applying protection to ideas that would be developed anyway is a net loss (it slows, not speeds technical advances) and thus is not intended by patent law. For more background and analysis, check out the League for Programming Freedom's web site, and remember---you can't own an idea, and you can't even limit its use unless society thinks it'll get something back for allowing such limits.
The questions above are on target, and here's one I haven't seen yet: How long will it take scum to poke holes in Google's algorithm, the way sites try to spam Altavista et al., by mindlessly repeating words in the background or title?
The old CPU network is a danger, but can be mitigated by continuing to use the older search engines. (Perhaps Google could add an `old-style search' option?) So long as we use other methods as well, and add the (often serendipitous) results to our links, new sites will join the crowd. If we use only one engine to the exclusion of others, though, the results base will stagnate. Google's great, but can't be the only one.
On the other hand, it will take (significantly?) more effort to spam Google, particularly if they deprecate links between pages in the same site/subdomain. It'll take a conspiracy of spammers, or the effort to set up multiple domains, to generate misleading results. Many small fry just won't bother.
That said, Google can return some funky finds, too. I was looking into a rumor that Volvo's engines for the 240 series were originally designed for farm tractors, and ended up at a site explaining ``You know you're an old-tractor freak when...''---all is not lost.
...couldn't we harness the/. effect to get them money? I'll send 'em $100 for a 1e-6 or so stake in the project. (Modify the factor for the amount of money desired---if they sold 49.9999...% of Google at that factor, they'd get $50M if fully subscribed.)
Slashdot Mirror
In a former life, I was an co-op student at Western Union (Anyone remember them? They used to actually transfer data as well as money.), and I was involved in leasing data lines from AT&T and the local telcos wherever our customers needed connections.
Some of the larger outfits were indeed interested in redundancy and paid a premium for, say, two links from NYC to SF, one by way of Chicago, and another through Houston. We frequently had trouble verifying that we were really getting independent routes.
The telcos bundled connectivity (back when a single voice line was lotsa data links [after they'd been digitized]) in a hierarchy so deep that it took days or weeks to verify that every link in the route used facilities physically separate from every other.
Why? 'Cause to Ma Bell, bandwidth is fungible. Got noise on link 37A from Manhattan to Albany? Take it out of service for maintenance, and swap in some spare bps from Manhattan to Jersey City to Albany. It's all the same....
Until a manhole floods in Jersey City, and *oops* it turns out your route to Chicago (formerly via Albany) is now cheek-by-jowl with the wire to Houston, and you're dead!
Among other things, we supposedly rented circuits with huge Do Not Reroute tags hanging all over them, but on occasion someone overlooked it, or worse, the facility two levels up the hierarchy got rerouted and our link went along for the ride unknowingly.
I wish them well---it can only be much worse with fiber optics these days.
I'm worried about the rush to fund Internet information sources from advertising---it'll end up heading the same way TV has: it's a medium to sell consumers to producers, with the content a distant second, useful only as a lure for the consumers.
Micro-payments would be an ideal way to avoid this problem, but the mechanisms to ensure reliability and anonymity, though extant, are pitifully slow in being implemented, and are further hindered by the absence of standards. I'd gladly pay, say, USD 0.0001 (a hundredth of a cent for the math-impaired) for any web page I'm served, so long as it's totally transparent to me. There are plenty of sites out there that get much less than that (read: zero). Unfortunately that's not going to happen soon.
So, how about those sites to whom I'd pay significantly more than that, maybe even an order of magnitude more, like Slashdot?
Anonymous micro-payments are good for surfing, but when I've found someplace really worthwhile, why shouldn't it offer advertising-free pages to me in exchange for cash? Are you listening, Rob?
How much do you get in advertising revenue for my presence? I bet I'd pay much more than that to rid my pages of ads. All you need to do is offer a, say USD 5.00--10.00 annual subscription (is this in the right ballpark?) to volunteers. I give you a credit-card number, you bill me annually, and when I log on, voila---no ads. If enough think this is worthwhile, you could get rid of the ads for everyone. I wouldn't mind subsidizing a few dozen ACs and free riders to get a /. which couldn't care less about the feelings of anyone other than its readers.
Of course, this raises the specter of /. dominated by the paying customers, but keep the price down, and anybody can play. Not too bad!
Thanks for a great article. Tell Jon Katz to reflect for a while on what makes this one so good :-).
The registration price list shows prices for registration ``before June 31'' and ``after June 31''---so we have 24 hours of undefined prices.
(Yes, I know about simulators, but it's still not like rolling off a log.)
The E-biomed flap is highlighting a lack in U.S. (at least) society's educational/social system. We try to teach our children all manner of facts and algorithms, but I've never seen (short of college) an explicit attempt to teach the need to evaluate information, and more importantly the methods for such evaluation. As the state of the nation demonstrates, we desperately need such.
The rise of the Internet has only exacerbated a pre-existing problem, as the existance of sleazy politicians, Ponzi schemes, and homeopathic medicine amply demonstrate. Until we get up off our arses and start looking into the information we receive and the biases of those supplying it, we will be putty in the hands of scum.
But, we can't check everything ourselves, we must delegate some of the examination to others, and the question we're missing is: ``How do we evaluate the trustworthiness of an informant?''
Such an evaluation is a case of diminishing recursion: Those we already trust can point us to informants they trust, but if we're wise we invest less trust the further the chain extends. But whom do we trust de novo? I claim that the only way to develop trust is through memory of previous information that has been checked afterward.
Did ``trickle-down'' economics work? Ask yourself how your situation changed during the Reagan-Bush years, then apply what you've learned to the trustworthiness rating of those supporting it. Does /. supply unerring information? Check up on the items you're interested in and adjust your trust accordingly. Without the effort to remember and evaluate, we're no better than animated bags of prejudice.
Now, down to cases: Why do I trust the New York Times reporting more than that of the National Enquirer? Because they've been proven to be full of malarkey less often. (Too frequently, but still less often.) Why do I trust the Merck Manual more than the local herbalist's gazette? Because the Merck tries to base itself on replicable data instead of anecdote. Does that mean one's gospel and the other's fish-wrapping? No---trust is not a binary value. It runs the gamut from nearly zero (but, if a fool says the Sun rises in the East, it's no less true) to almost complete (but even I'm fallible).
We need to start teaching trust evaluation. One of the major problems in this day and age is the value placed on quick information. Evaluation and verification takes time, and information immediately available is inherently less trustworthy---something many on /. ignore. For most information, speed is not of the essence, accuracy is. The only way we can trust speedy information is if we trust the informant, and in e-info, the informant is all too often someone we've never heard of. If Alan Cox reports a kernel bug that can wipe my hard drive, I'll act on it a lot faster than if it comes from a /. AC.
But, trust also has subject limits. I'll listen respectfully to Mr. Cox on the kernel, programming, etc., but when it comes to investment information or medical treatments he's just another tyro until I've evaluated his abilities in those fields.
[This is one reason I'm happy that we have accounts available on /.---I certainly don't know who really signed up for the LTorvalds account (if such exists), but if an account holder is reasonably careful with the password, I can trust that the DonkPunch I heard from last week is the one posting today, and can apply the trust rating I previously developed for that poster to the current post.]
Thus, information evaluation is based on how long the info's been available for evaluation, and the track record of those who've evaluated it (if I'm unable to do so personally). Teach that to society, and we'll all live better (except for the scoundrels out there).
As for Dr. Varmus, it indeed sounds as if he's got his head screwed on straight. He's setting up a system where information is explicitly separated by speed of availability and depth of examination---take whichever you like, but remember to apply the appropriate assessment based on where you got it.
As people begin to pay more attention to reliability of information, we'll see more such web sites develop---they may not have the rumors that came through in the last 30 seconds, but they'll be the places you go when it really matters.
After reading the decision on the Bernstein case (see the /. article for a refresher), the typo in Stephenson's book struck me as a perfect defense.
The decision refers to the Government's claim that the functional aspects of crypto override any First-Amendment issues. (See pp. 4235, 4236, 4238n., and the dissent on pp. 4246ff.) Thus, we can export the first edition of Cryptonomicon in machine-readable form, free of obnoxious restraints, precisely because it doesn't function. A comment on the code would be nice, or a pointer to the errata, but if the code don't work, the Government can't claim it's crypto!
Any takers? :-)
[Found in a university term paper analyzing Kubrick's A Clockwork Orange.]
Would it be worth a try to let readers nominate moderators? That way, when I notice a poster whose comments are really cogent, I can mark that person as a potential moderator.
Of course, this posits that good comments come from people with the balance and judgment to moderate well. I think that's true as a first approximation.
How to use this input, however? Count up the number of times a particular poster gets nominated, and the highest numbers get moderator privileges? (`Nomination' is really voting.) Alternatively, if the count goes over some threshold, have a human look at that person's posts and decide yea or nay. (`Nomination' is just that, with other criteria for acceptance.)
I suggest giving the first mechanism a try, because there are some folks whose judgment I respect, and would like to see have some input based on their abilities rather than pure numerology. (<nose color="brown">Not that the numerology isn't a damned good shot at automating a tough call.</nose>)
Maybe nominations can come only from readers in domains other than the nominee's? Perhaps only from the older 2/3 of the population? (Both of these to prevent someone from registering an bunch of IDs and performing auto-nomination.)
That's `cause he owns it. I've read that Rick Sharp (head of Circuit City) thought up the outlines of the product, then assigned some good folks to flesh it out. I don't think he's going to fire himself. :-) (Hi, Rick!)
(Dupe so I'm logged in---posting code forgot between preview and post last time.)
It's a perfectly good English interpretation to read your intro as ``[Christiansen] argues, with the FSF and RMS, that...'' (note the added commas), which means he agrees with them. In fact, that's what I expected to read, and was taken aback when the article said the exact opposite. :-)
English is wonderful, not least due to its openness to the well-place ambiguity, but when you're writing blurbs, make sure there are none you didn't intend.
</nitpick>
This is a copyright suit, not a patent one. Prior art invalidates a patent, but if you have a copyright, you have rights to the expression (roughly, the way you coded it), not to the algorithm.
Have you ever been within smelling distance of a lawsuit? If not, read, for example, coverage of the Microsoft suit the DoJ's prosecuting.
I have a number of friends who are lawyers (and actually have working consciences!), and get the inside scoop from them about the operation of the law. As P. J. Plauger observed in an Embedded Systems Programming column, anytime you go to court, you've lost, regardless of the merits of your case.
If we accept licenses that allow courts to decide what the license language means (as a primary function, not just as secondary to deciding whether the license has been violated), we're signing our own death warrant. Anyone with a few bucks to rub together can bring all development to a complete halt at any time for any project using such a license.
It's time to go back to basic sources. The United States Constitution, Article I, Section 8, states:
(Emphasis added---see your local social contract for other countries.)It's not about an inventor's ownership---there is no ownership of ideas, nor even any limits on use except as society has decided it's to its advantage to allow such.
The advantage posited is that if society offers remuneration to those who come up with new ideas, we'll get more such ideas, and civilization will advance that much more rapidly. (I'm all in favor of penicillin and contact lenses, myself.) The remuneration is in the form of a monopoly on use of the idea (for a limited time), allowing the thinker to make money on it before others get a shot at copying it.
This was a Good Thing for ideas that took years of sweat, tons of metal, and a lot of limited and expensive resources to develop, like, say, Bessemer furnaces, safety pins, and leading-shoe brakes. If the idea wouldn't be implemented without a pay-back for the thinker, society was less likely to reap its benefits. So, we say (for patents) ``Here, tell us what your idea is, document it for all to see, and in return for that openness we'll help you make money from it---not because you have any ownership of an idea, but because we'll get the benefit of having it sooner, and letting other thinkers build on it.''
The vital concept here is that patents are for the good of society, not the person with the idea. If an idea is most likely going to be developed anyway, there's no social benefit to limit use of it. Further, software is one of the most likely areas for ideas to be developed without monopoly, simply because coding is cheap, attracts thousands (millions?) of people, and is the most fun you can have with your clothes on. If one hacker doesn't generate a given idea, there'll be another one along in a minute.
Unfortunately, over the decades, people got (remained?) greedy, and specialized enough to have never studied their own history, and so began to think that an idea, of all things, could have an owner, and that it was their right to have a monopoly and control access to the idea.
Look back on the development of software: in the past 60 years is there any evidence that its progress has been hampered by people afraid to develop ideas for lack of protection? Hah! Never in the history of technology has a field advanced so rapidly or been so fecund, and most of it has been due to the open availability of ideas. Until the 1980s no one could patent software, and copyright seemed to work quite nicely, thank you. Now, unfortunately, we have people who think that because they spent a year or two on an idea, they should be allowed to prohibit anyone else from using the idea, even though it's hardly novel, or even notably hard---tedious in reducing to a useful form, but no more.
So, someone building a virtual machine (which IBM made mega-bucks on in the '60s and '70s [and may be still, for all I know]) cannot be allowed to say ``I'm the only person who can use this idea for the next 17 years'' (or is it up to 20 now?) when the same idea has occurrred to anyone who's written an emulator.
Applying protection to ideas that would be developed anyway is a net loss (it slows, not speeds technical advances) and thus is not intended by patent law. For more background and analysis, check out the League for Programming Freedom's web site, and remember---you can't own an idea, and you can't even limit its use unless society thinks it'll get something back for allowing such limits.
Check out the copyright on the articles---you can copy them freely---O'Reilly doesn't own them, and the authors allow copying.
At least in the copy of the NT 4.0 EULA I have handy, the words are:
(Emphasis added.)
So, don't go to MS, go to wherever you bought the system from.
MS never wants to deal with you, only with your money.
Now let's see...How long would it take to download 8/1000 of a bit---and how do you split bits anyway?
Oh, he meant 10 Mb? Never mind...
(Please pardon the momentary humor-impairment.)
The questions above are on target, and here's one I haven't seen yet: How long will it take scum to poke holes in Google's algorithm, the way sites try to spam Altavista et al., by mindlessly repeating words in the background or title?
The old CPU network is a danger, but can be mitigated by continuing to use the older search engines. (Perhaps Google could add an `old-style search' option?) So long as we use other methods as well, and add the (often serendipitous) results to our links, new sites will join the crowd. If we use only one engine to the exclusion of others, though, the results base will stagnate. Google's great, but can't be the only one.
On the other hand, it will take (significantly?) more effort to spam Google, particularly if they deprecate links between pages in the same site/subdomain. It'll take a conspiracy of spammers, or the effort to set up multiple domains, to generate misleading results. Many small fry just won't bother.
That said, Google can return some funky finds, too. I was looking into a rumor that Volvo's engines for the 240 series were originally designed for farm tractors, and ended up at a site explaining ``You know you're an old-tractor freak when...''---all is not lost.
...couldn't we harness the /. effect to get them money? I'll send 'em $100 for a 1e-6 or so stake in the project. (Modify the factor for the amount of money desired---if they sold 49.9999...% of Google at that factor, they'd get $50M if fully subscribed.)