A Federal appeals court just ruled in the Padilla case that it's ok for the President and his forces to declare an American citizen to be an "enemy combatant" and hold him indefinitely without trial. It's still possible for the Supreme Court to reject that and affirm that we do have fundamental rights, but Bush's nominee for Chief Justice, John Roberts, has previously ruled in favor of the Administration on several similar cases. It's possible that he'll nominate Alberto Gonzales, the White House lawyer who believes that anything done by American forces isn't torture so whatever they were doing to prisoners in Gitmo wasn't illegal - and the main opposition to him are right-wingers who think he isn't "conservative" enough.
I thought their choice of "Four Pentium M chips at 1.7GHz" sounded more absurd than "One 6.8 GHz Magic Vaporware Chip" - 6.8==4*1.7, but I don't know why you'd need that many CPU chips burning batteries, especially if you're claiming whatever reasonably-long battery life they claimed.
Also, remember that you really can't compare processors based on raw clock rate - it's like comparing internal combustion engines based on RPM, rather than horsepower/torque/etc. It's quite possible that their 6.8 GHz could be the speed of a DSP, or adding up all the parallel processor elements of a vector processor or something similarly bogus.
On the other hand, because they've developed a magic new vaporware-based memory technology, it may be that 256MB of cache is a perfectly reasonable thing to include in their processor. If *I* were designing a laptop today, using normal technology, I'd almost certainly want to include a gigabyte or two of flash RAM as a disk cache - it lets you install the OS and some of your main applications in it, so you don't need to power rotating machinery most of the time.
They're not using hard disk, but they're also not using flash memory. They're using their own special technology, which is an order of magnitude cheaper than flash (they say $6/GB, compared to flash prices which tend to be about $50/GB), which is where the $6000 for 1TB figure came from.
Think of it as Vapor-State drives - the chances that one small company that nobody's ever heard of has simultaneously developed a radically new memory technology and also a general-purpose CPU that's significantly faster than Intel's hottest laptop or desktop CPU sounds pretty minimal.
When Microsoft charges money for software or features, everybody bitches because they're a greedy monopoly. But when MS includes features for free, everybody bitches and sues them, because they're evil greedy monopolists trying to undercut their competition.
The Eurocrats recently forced them to release a version with no media player on it, which everybody viewed as silly and ignored. But earlier, when they gave away IE for free, Netscape got the Feds to investigate them, because MS was greedily interfering with Netscape's business model of giving away browsers for free...
Tylenol's going to do a lot more damage if you're taking it at Vicodin-abuser levels rather than overdose levels, trashing your liver and kidneys. Aspirin can rot out your stomach and give you ulcers and such, but it's an acute problem that gets your attention more directly. (On the other hand, I suppose if you're taking opiate painkillers, you might not notice the stomach pain as quickly.)
Those 72000 prescriptions were probably for far fewer than 72000 customers.
If you're trying to buy Rush Limbaugh quantities of painkillers, and the first bottle or two from Spamboy arrive (and the price is reasonable compared to other available sources), you're probably going to buy a lot from the same source - much easier than shopping around, and the fact that Spamboy is clearly running a sleazy operation means that he's not going to check the quantities you're buying as carefully as your neighborhood pharmacist would.
Also, if his price is reasonable *enough*, some of his customers are probably buying wholesale quantities from him and retailing to their own customers.
Different people react differently to different drugs, and while it's well-known that it's true for opiates, it's also true for novocaine. I once had novocaine not work at the dentist, and he said sometimes the stuff doesn't work if you haven't eaten, and sent me out to go have lunch and come back in the afternoon; after awkwardly negotiating lunch with a numb mouth (:-) I came back and the stuff worked fine.
Some of my friends find codeine makes them hallucinate or feel really bad. The one time I took Percocet I felt awful, though it was just after a root canal which makes it a bit hard to compare. Codeine doesn't bother me, and generally helps pain. Think I'll go take one.
Some people do RAID for reliability; others do it for speed. If your real goal is speed, then you either need to get somebody to develop a 15000 RPM laptop disk drive, which will probably have less capacity and put out more heat and may very well use more battery than two conventional drives, or else you need to do RAID. It doesn't have to be hardware RAID - software RAID is more flexible and you've almost certainly got the spare CPU to do it - but if you're selling chipsets, it's not that much trouble to add the basic RAID functions.
On the other hand, if your goal is more capacity, you may very well want two disks, but that doesn't mean they need to be RAID. Sometimes it can help (duplicating critical file systems, etc.), though some operating systems from the northwest side of North America aren't very good at giving you control over what resources really live where except for some of your user directories.
With either goal, power-management systems can do reasonably intelligent things with controlling the two disks. For instance, if you're not reading the disks very often, it's easy to scale back to only reading from one drive, and even if you're writing, you could do a sloppy-mirroring RAID approach that buffers writes to the second drive if it's currently spun down.
SpamAssassin Tools - AK-47s, Knives, or Nukes?
on
Ask Jonathan Zdziarski
·
· Score: 2, Interesting
So when you're trying to assassinate spammers, do you favor precisely targeted tools like knives, medium-scale tools like AK-47s, or nuke-them-from-orbit solutions?
I guess the more serious version of this question is the tradeoff of precision and false negatives vs. overkill and false-positives. For instance, my email provider lets me pick country-blacklists, so I reject all email from China, Korea, and Nigeria, where I don't know anybody, and Japan gets accepted with extra filtering, because I know a couple people there who normally don't send me mail - it's not quite a nuke-Asia-from-orbit approach, because people who actually do want mail from people in China can accept it, but people who don't can reject it all and lose the occasional message from a friend at a cybercafe.
Sure, Moore's initial observation was about transistors, but the important core of the observation was that capacity or performance was doubling every N months. Other than chip-design junkies, nobody really cares how many transistors you have - we care how much memory you can get, or how fast a CPU can crunch numbers, or how much you have to pay for the same amount of capacity - and *those* things have improved roughly exponentially over the last ~20 years. The growth has been driven by the market's ability to invest in technology development, which is partly driven by the software and applications market's ability to use whatever hardware capacity you can afford to buy, either for productive use or for flashy decoration that makes your machine *look* better than your competitors'. (And by games, which have been a critical driver the last few years.)
A 12-month version of the statement was somebody at Sun saying that the speed of a Sun computer in MIPS = 2**(year-1985), which was pretty much true for a couple of years. MacNealy or somebody described their early architecture as "3M - 1 Megapixel display, 1 MIPS CPU, 1 Megabyte of memory" (probably Sun-1.) The Sun 2/50 in my attic has an 1152x900 screen. The medium-range IBM laptop I'm typing this on has about 1000 times faster CPU, 512x as much memory, 60GB more disk (2/50 was diskless:-), and the screen may be 32-bit color and radically accelerated instead of black&white, but it's still only 1024x768:-(....
This isn't a single chip - it's a range of chips with widely differing specifications and capabilities, or at least it would be if they were actually giving out any real specifications which all the articles were a bit sparse in. Basically, Steve Jobs is getting the use of Intel's chips, and Intel is getting to borrow the Reality Distortion Field so they can talk about how insanely great the new chips are going to be when they arrive (though it's kind of surprising to get this level of fluffy marketing talk when they're also showing that real silicon exists.)
They've got some common design elements, and different tradeoffs of number of processor cells vs. power control, etc., but the big advantage of commonality is better chipset support. It costs a lot of money and design-cycle time to have to crank out too many different chipset designs, and simplifying them makes it easier for motherboard manufacturers to support their chips over a longer period of time at lower costs.
Sure, battery capacity per volume|price|weight hasn't been quadrupling every year the way disk drives or RAM were for a while, but they've still been on a pretty steep technology curve compared to most non-computer things you use. Lithium batteries do a lot better for most applications than NiMH, and NiMH were a lot better for most applications than NiCd, and most of the newer battery chemistries may not be totally clean but they're a lot less nasty than NiCd was.
The problem is that, while batteries really have been improving a lot, most laptop makers think that customers want the fastest and most powerful laptops they can get, with the biggest and brightest possible screens on them, as lightweight and thin as they can get, so the power demands of laptop computers have been sucking down every bit of power the batteries can deliver, and since power demand is always going to exceed supply, your batteries are going to feel wimpy because the manufacturers are going to aim for a 1-2 hour battery life because that's the shortest that customers will accept (and of course, a year later your battery life will probably have degraded to half of that unless you're strictly using it as a desktop most of the time. Batteries are the bottleneck because *something* has to be, and that's what the market seems to care about least.
If you wanted to build a Transmeta or Via low-power slow-but-adequate machine with a smaller less bright screen, and were willing to run lower-horsepower applications (Linux/BSD/olderWindows) and not play Gamez on the machine, you'd be surprised what kind of battery life you could get with modern batteries, especially if you start using flash for your main drives instead of rotating platters. Apple seems to be the only manufacturer out there who's emphasizing battery life on anything except niche machines, and they've done a few that claim to support the "airplane across North America or the Atlantic" market.
Of course they wouldn't abuse those powers. That would be *impolite*, and and they're *Canadians*.
Now, if they were actually carrying out the wiretaps on behalf of US police/thugs/spooks/politicians/military/etc, that's a different game entirely, because then you're looking at the lowest-common-denominator of those groups' values, and what the [expletive deleted] is the problem with that?
More precisely, you use a Diffie-Hellman key exchange to create a one-use session key, and you use your public/private keypair to *sign* the keyparts, which prevents man-in-the-middle attacks and some other kinds of forgery. Even if the police or courts force you or your ISP to give up information about your keys or email, it doesn't do any good except to make it possible to impersonate you in the future, and it's reasonable to argue in court that they shouldn't be able to do that.
It's also reasonable to argue that they shouldn't be eavesdropping on you, especially without a court order, but it's a lot harder, and that doesn't prevent them from collecting whatever information your ISP has or forcing your ISP to give them copies of any of your port 25 traffic in the future.
You saw that NetBSD-based toaster at Linuxworld, didn't you?
A kilowatt is a bit light-weight for a toaster, but on the other hand it doesn't need highly filtered DC in several different voltages, so the power supply can look suspiciously like the power cord used by other power supplies...
Certainly *some* people in the US government are all in favor of *some* videogames increasing American youth's aggressive behavior, interest in violence, alignment with one side in conflicts and belief that the other side is evil and should be killed....
I think New Jersey has laws against card-counting, though it's been a while since I lived there. It's definitely legal for casinos to kick people out of public establishments for playing rigged games in ways that the people who rigged them don't like, and they've got laws that permit them to do it - and *that's* cheating. Standard Blackjack rules make it possible for good players to occasionally beat the house systematically, and they've already made a number of changes to the game to make that harder (more decks of cards in the shoe, etc.) (as well as making it harder for dealers to cheat the house), but kicking out player when they're winning is poor sportsmanship. They ought to either solve the problem by making more player-neutral rules, or solve it by giving more free drinks to players who appear to be winning too consistently (or by letting them win, and giving more free drinks to the 4-5 suckers who are losing money at the table that the winner is at.)
If you're just spending one weekend in Vegas, and you get kicked out of one casino because you're actually playing to win, yes, you can walk down the street and play at another casino (at least if it's not run by the same company, which several sets of them are), and worst case is you might have to drive out to Sam's Town or some other off-strip location.
But if you're playing blackjack as a *business*, you need to be able to keep playing. Casinos do talk to each other about problem players, and while they're more concerned about actual cheaters (dealers in league with players, counterfeit chips, etc.), if they're seeing the same players winning too often at games that are rigged in favor of the house, they're going to keep track of who's doing it and stop them. And it's usually groups of players, not just individuals - making money off card counting is usually a team sport, with division of labor between the different players to try to maximize information collection and exploitation while reducing visibility. So you may have some of your players betting at lower levels and doing the counting while another player does the high-roller bit on the tables that have the right odds.
Claiming "SMTP is Broken" without any better ideas
on
Ending Spam
·
· Score: 1
I'm tired of hearing people rant that "We have spam because SMTP is Broken, and SOMEBODY ought to fix it", when they don't really have any better ideas. If you've got any sense of history, you'd remember the complexity of X.400 (which has a lot to do with why almost nobody uses it), and they'd remember the newer UUCP versions that had authentication built in (doesn't stop spam either), and relatively closed systems (market forces either killed them or forced them to interface with Internet mail.)
The fundamental problem is that technology pushed the *costs* of sending mail and creating identifiers (IP addresses, domain names, email addrs, etc.) to near-zero and the cost of finding recipients to near zero, human nature makes it profitable to send gullible people mail if you've got no morals, and the popularity of the internet means that people with no morals can easily get the tools to use it. Willingness to spam is a social problem, and economics have made it possible to become an actual problem. The real cost of sending mail isn't likely to go up (encryption affects it a bit, but CPU time is basically free, or you can attempt to impose artificial prices on email transmission (which will fail, if you get it accepted at all, because they don't match real prices.) You can use technology to increase the cost of discovering recipients, using things like tagged addresses and subdomain-per-user naming that increase the search space, and you can use technology to reduce the amount of mail a given group of senders can send to a given receiver. *Recipients* can impose prices or other throttling mechanisms on senders without disrupting most of the other infrastructure, which can help - I know a number of people who find that simple TMDA/Captcha techniques kill off most of their spam, by increasing the cost of discovering an email address that they'll *read* (the cost is the attention spam of having a real human read the captcha image, plus the need to use a real email address to send from instead of a bogus one) - but even they say that it annoys some people they'd really like to get email from.
Sure, some details will change, and spammers and anti-spammers will pick up new tricks and abandon old ones, and the percentages of email that are spam will keep changing (normally up, but I saw one recent article saying it had dropped significantly in the last year.) But most of the fundamentals don't change much, or at least not very fast. Filtering techniques, Bayesian analysis, collaborative filtering, etc. are a solid core of knowledge that will continue to be useful.
Rule 1 (Spammers always lie) won't change, though occasionally they'll think of new things to lie about. Rule 2 (Spammers are Stupid) won't change, though of course some spammers violate this rule, and some spammers can hire smart people to work for them, and enough of them are sufficiently persistent skr1pt k1dd13z that it sometimes makes up for stupidity.
The latest and greatest spam-blocking technique will last a while before spammers find a way around it - it's somewhat of a losing game, because if it works well enough to be widely popular, it becomes a target for spammers to work around, though if it's effective and obscure, it'll work for you and your friends for a lot longer.
PC users will continue to run insecure operating systems without administering them well, so there'll always be zombies for spammers to abuse. Windows automatic updates will gradually help this, but not only will new OS bugs get discovered frequently, but users will insist on running trojan horses that pretend to be new amusing programs, breaking any semblance of security.
Fingerprint readers and other biometric sensors are almost always a misguided idea, often an evil one, and generally not implemented well. You could get much more useful capabilities by including a small keypad on it, which could be used for passwords if you need them (which you sometimes do, depending on your application), and maybe a little 1-or-2-line LCD display for status.
Rule 1: Spammers always lie. Rule 2: Spammers are stupid.
Rule 2 means that spammers usually aren't competent, but if one of them is, he'll set up a corporation to isolate his activities from the spamming, so if anybody bothers to go to the effort to prosecute for spamming, it's the corporation that gets busted and not the individual, and it's the corporation's assets that are exposed to the court, not the actual spammer's. Of course, the corporation will make sure to spend its money paying employees or buying services from people or whatever, so the only assets are a $100 corporate charter, some petty cash, and maybe a couple of cheap PCs in a colo center (but they'd probably be leased also, so if the spammer's corporation gets busted, they just stop paying their PC rent.)
It's hard enough to prosecute anybody for spamming, and it's a lot more work to pierce the corporate veil and bust the individuals responsible for it - they'll do that for a multi-billion-dollar Enron or Worldcom scam, but a bottom-feeding spammer in a mobile home park isn't worth the effort, and a medium-sized million-dollar spammer can go to a bit more work and use a $1000 off-shore tax-haven corporation instead of a $100 Delaware corporation.
I *have* tracked down at least one spammer to a mailbox at the address of "The Company Corporation" in Delaware - they're the canonical $100 Delaware Corporation provider, so it was real obvious that there'd be no use chasing them further.
This isn't a criminal prosecution, in which the government would be trying to prove that the accused did something criminal (Eliott Spitzer was also doing that, but this is a different case.) This is a lawsuit, in which Microsoft was trying to prove that Scotty did harm to Microsoft, and should be required to compensate Microsoft for that harm. The issue of "Settling" here is whether the parties involves go to the time and expense of finishing a trial, where the amount of money that gets paid is much less predictable and therefore riskier to both sides, as opposed to deciding on a specific amount of money that's less than the defendant might have had to pay if he'd lost badly but more than he'd have to pay if he won or didn't lose too badly. Also, for this case, Scotty was trying to get off the hook by filing bankruptcy to avoid the judgement, and this gets around that problem.
I really liked the early 419 scams, where the perp was claiming to be some corrupt official trying to illegally ship stolen money around, because anybody who fell for them was themselves corrupt and greedy and couldn't go to the cops because they were participating in what they *thought* was a criminal activity, though in reality they were the victim as well as a wannabee perp. Too bad the rest of us have to be inundated with spam in the process.
The newer ones are too tame - the fake lotto scams exploit stupid greedy people (but so do the government-run lotteries they're usually pretending to compete with), and the "dying cancer patient wants to do something good with the rest of her life" tearjerkers are really only exploiting the gullible, who don't deserve to be abused the way the classic 419 victims do.
Slashdot Mirror
A Federal appeals court just ruled in the Padilla case that it's ok for the President and his forces to declare an American citizen to be an "enemy combatant" and hold him indefinitely without trial. It's still possible for the Supreme Court to reject that and affirm that we do have fundamental rights, but Bush's nominee for Chief Justice, John Roberts, has previously ruled in favor of the Administration on several similar cases. It's possible that he'll nominate Alberto Gonzales, the White House lawyer who believes that anything done by American forces isn't torture so whatever they were doing to prisoners in Gitmo wasn't illegal - and the main opposition to him are right-wingers who think he isn't "conservative" enough.
Also, remember that you really can't compare processors based on raw clock rate - it's like comparing internal combustion engines based on RPM, rather than horsepower/torque/etc. It's quite possible that their 6.8 GHz could be the speed of a DSP, or adding up all the parallel processor elements of a vector processor or something similarly bogus.
On the other hand, because they've developed a magic new vaporware-based memory technology, it may be that 256MB of cache is a perfectly reasonable thing to include in their processor. If *I* were designing a laptop today, using normal technology, I'd almost certainly want to include a gigabyte or two of flash RAM as a disk cache - it lets you install the OS and some of your main applications in it, so you don't need to power rotating machinery most of the time.
Think of it as Vapor-State drives - the chances that one small company that nobody's ever heard of has simultaneously developed a radically new memory technology and also a general-purpose CPU that's significantly faster than Intel's hottest laptop or desktop CPU sounds pretty minimal.
The Eurocrats recently forced them to release a version with no media player on it, which everybody viewed as silly and ignored. But earlier, when they gave away IE for free, Netscape got the Feds to investigate them, because MS was greedily interfering with Netscape's business model of giving away browsers for free...
Tylenol's going to do a lot more damage if you're taking it at Vicodin-abuser levels rather than overdose levels, trashing your liver and kidneys. Aspirin can rot out your stomach and give you ulcers and such, but it's an acute problem that gets your attention more directly. (On the other hand, I suppose if you're taking opiate painkillers, you might not notice the stomach pain as quickly.)
If you're trying to buy Rush Limbaugh quantities of painkillers, and the first bottle or two from Spamboy arrive (and the price is reasonable compared to other available sources), you're probably going to buy a lot from the same source - much easier than shopping around, and the fact that Spamboy is clearly running a sleazy operation means that he's not going to check the quantities you're buying as carefully as your neighborhood pharmacist would.
Also, if his price is reasonable *enough*, some of his customers are probably buying wholesale quantities from him and retailing to their own customers.
I once had novocaine not work at the dentist, and he said sometimes the stuff doesn't work if you haven't eaten, and sent me out to go have lunch and come back in the afternoon; after awkwardly negotiating lunch with a numb mouth (:-) I came back and the stuff worked fine.
Some of my friends find codeine makes them hallucinate or feel really bad. The one time I took Percocet I felt awful, though it was just after a root canal which makes it a bit hard to compare. Codeine doesn't bother me, and generally helps pain. Think I'll go take one.
On the other hand, if your goal is more capacity, you may very well want two disks, but that doesn't mean they need to be RAID. Sometimes it can help (duplicating critical file systems, etc.), though some operating systems from the northwest side of North America aren't very good at giving you control over what resources really live where except for some of your user directories.
With either goal, power-management systems can do reasonably intelligent things with controlling the two disks. For instance, if you're not reading the disks very often, it's easy to scale back to only reading from one drive, and even if you're writing, you could do a sloppy-mirroring RAID approach that buffers writes to the second drive if it's currently spun down.
I guess the more serious version of this question is the tradeoff of precision and false negatives vs. overkill and false-positives. For instance, my email provider lets me pick country-blacklists, so I reject all email from China, Korea, and Nigeria, where I don't know anybody, and Japan gets accepted with extra filtering, because I know a couple people there who normally don't send me mail - it's not quite a nuke-Asia-from-orbit approach, because people who actually do want mail from people in China can accept it, but people who don't can reject it all and lose the occasional message from a friend at a cybercafe.
A 12-month version of the statement was somebody at Sun saying that the speed of a Sun computer in MIPS = 2**(year-1985), which was pretty much true for a couple of years. MacNealy or somebody described their early architecture as "3M - 1 Megapixel display, 1 MIPS CPU, 1 Megabyte of memory" (probably Sun-1.) The Sun 2/50 in my attic has an 1152x900 screen. The medium-range IBM laptop I'm typing this on has about 1000 times faster CPU, 512x as much memory, 60GB more disk (2/50 was diskless :-), and the screen may be 32-bit color and radically accelerated instead of black&white, but it's still only 1024x768 :-(....
They've got some common design elements, and different tradeoffs of number of processor cells vs. power control, etc., but the big advantage of commonality is better chipset support. It costs a lot of money and design-cycle time to have to crank out too many different chipset designs, and simplifying them makes it easier for motherboard manufacturers to support their chips over a longer period of time at lower costs.
The problem is that, while batteries really have been improving a lot, most laptop makers think that customers want the fastest and most powerful laptops they can get, with the biggest and brightest possible screens on them, as lightweight and thin as they can get, so the power demands of laptop computers have been sucking down every bit of power the batteries can deliver, and since power demand is always going to exceed supply, your batteries are going to feel wimpy because the manufacturers are going to aim for a 1-2 hour battery life because that's the shortest that customers will accept (and of course, a year later your battery life will probably have degraded to half of that unless you're strictly using it as a desktop most of the time. Batteries are the bottleneck because *something* has to be, and that's what the market seems to care about least.
If you wanted to build a Transmeta or Via low-power slow-but-adequate machine with a smaller less bright screen, and were willing to run lower-horsepower applications (Linux/BSD/olderWindows) and not play Gamez on the machine, you'd be surprised what kind of battery life you could get with modern batteries, especially if you start using flash for your main drives instead of rotating platters. Apple seems to be the only manufacturer out there who's emphasizing battery life on anything except niche machines, and they've done a few that claim to support the "airplane across North America or the Atlantic" market.
Now, if they were actually carrying out the wiretaps on behalf of US police/thugs/spooks/politicians/military/etc, that's a different game entirely, because then you're looking at the lowest-common-denominator of those groups' values, and what the [expletive deleted] is the problem with that?
It's also reasonable to argue that they shouldn't be eavesdropping on you, especially without a court order, but it's a lot harder, and that doesn't prevent them from collecting whatever information your ISP has or forcing your ISP to give them copies of any of your port 25 traffic in the future.
A kilowatt is a bit light-weight for a toaster, but on the other hand it doesn't need highly filtered DC in several different voltages, so the power supply can look suspiciously like the power cord used by other power supplies...
Certainly *some* people in the US government are all in favor of *some* videogames increasing American youth's aggressive behavior, interest in violence, alignment with one side in conflicts and belief that the other side is evil and should be killed....
I think New Jersey has laws against card-counting, though it's been a while since I lived there. It's definitely legal for casinos to kick people out of public establishments for playing rigged games in ways that the people who rigged them don't like, and they've got laws that permit them to do it - and *that's* cheating. Standard Blackjack rules make it possible for good players to occasionally beat the house systematically, and they've already made a number of changes to the game to make that harder (more decks of cards in the shoe, etc.) (as well as making it harder for dealers to cheat the house), but kicking out player when they're winning is poor sportsmanship. They ought to either solve the problem by making more player-neutral rules, or solve it by giving more free drinks to players who appear to be winning too consistently (or by letting them win, and giving more free drinks to the 4-5 suckers who are losing money at the table that the winner is at.)
But if you're playing blackjack as a *business*, you need to be able to keep playing. Casinos do talk to each other about problem players, and while they're more concerned about actual cheaters (dealers in league with players, counterfeit chips, etc.), if they're seeing the same players winning too often at games that are rigged in favor of the house, they're going to keep track of who's doing it and stop them. And it's usually groups of players, not just individuals - making money off card counting is usually a team sport, with division of labor between the different players to try to maximize information collection and exploitation while reducing visibility. So you may have some of your players betting at lower levels and doing the counting while another player does the high-roller bit on the tables that have the right odds.
The fundamental problem is that technology pushed the *costs* of sending mail and creating identifiers (IP addresses, domain names, email addrs, etc.) to near-zero and the cost of finding recipients to near zero, human nature makes it profitable to send gullible people mail if you've got no morals, and the popularity of the internet means that people with no morals can easily get the tools to use it. Willingness to spam is a social problem, and economics have made it possible to become an actual problem. The real cost of sending mail isn't likely to go up (encryption affects it a bit, but CPU time is basically free, or you can attempt to impose artificial prices on email transmission (which will fail, if you get it accepted at all, because they don't match real prices.) You can use technology to increase the cost of discovering recipients, using things like tagged addresses and subdomain-per-user naming that increase the search space, and you can use technology to reduce the amount of mail a given group of senders can send to a given receiver. *Recipients* can impose prices or other throttling mechanisms on senders without disrupting most of the other infrastructure, which can help - I know a number of people who find that simple TMDA/Captcha techniques kill off most of their spam, by increasing the cost of discovering an email address that they'll *read* (the cost is the attention spam of having a real human read the captcha image, plus the need to use a real email address to send from instead of a bogus one) - but even they say that it annoys some people they'd really like to get email from.
Rule 1 (Spammers always lie) won't change, though occasionally they'll think of new things to lie about. Rule 2 (Spammers are Stupid) won't change, though of course some spammers violate this rule, and some spammers can hire smart people to work for them, and enough of them are sufficiently persistent skr1pt k1dd13z that it sometimes makes up for stupidity.
The latest and greatest spam-blocking technique will last a while before spammers find a way around it - it's somewhat of a losing game, because if it works well enough to be widely popular, it becomes a target for spammers to work around, though if it's effective and obscure, it'll work for you and your friends for a lot longer.
PC users will continue to run insecure operating systems without administering them well, so there'll always be zombies for spammers to abuse. Windows automatic updates will gradually help this, but not only will new OS bugs get discovered frequently, but users will insist on running trojan horses that pretend to be new amusing programs, breaking any semblance of security.
Fingerprint readers and other biometric sensors are almost always a misguided idea, often an evil one, and generally not implemented well. You could get much more useful capabilities by including a small keypad on it, which could be used for passwords if you need them (which you sometimes do, depending on your application), and maybe a little 1-or-2-line LCD display for status.
Rule 2 means that spammers usually aren't competent, but if one of them is, he'll set up a corporation to isolate his activities from the spamming, so if anybody bothers to go to the effort to prosecute for spamming, it's the corporation that gets busted and not the individual, and it's the corporation's assets that are exposed to the court, not the actual spammer's. Of course, the corporation will make sure to spend its money paying employees or buying services from people or whatever, so the only assets are a $100 corporate charter, some petty cash, and maybe a couple of cheap PCs in a colo center (but they'd probably be leased also, so if the spammer's corporation gets busted, they just stop paying their PC rent.)
It's hard enough to prosecute anybody for spamming, and it's a lot more work to pierce the corporate veil and bust the individuals responsible for it - they'll do that for a multi-billion-dollar Enron or Worldcom scam, but a bottom-feeding spammer in a mobile home park isn't worth the effort, and a medium-sized million-dollar spammer can go to a bit more work and use a $1000 off-shore tax-haven corporation instead of a $100 Delaware corporation.
I *have* tracked down at least one spammer to a mailbox at the address of "The Company Corporation" in Delaware - they're the canonical $100 Delaware Corporation provider, so it was real obvious that there'd be no use chasing them further.
This isn't a criminal prosecution, in which the government would be trying to prove that the accused did something criminal (Eliott Spitzer was also doing that, but this is a different case.) This is a lawsuit, in which Microsoft was trying to prove that Scotty did harm to Microsoft, and should be required to compensate Microsoft for that harm. The issue of "Settling" here is whether the parties involves go to the time and expense of finishing a trial, where the amount of money that gets paid is much less predictable and therefore riskier to both sides, as opposed to deciding on a specific amount of money that's less than the defendant might have had to pay if he'd lost badly but more than he'd have to pay if he won or didn't lose too badly. Also, for this case, Scotty was trying to get off the hook by filing bankruptcy to avoid the judgement, and this gets around that problem.
Sure, 186 mph is cool. But 186232 mph would really be a lot more fun...
The newer ones are too tame - the fake lotto scams exploit stupid greedy people (but so do the government-run lotteries they're usually pretending to compete with), and the "dying cancer patient wants to do something good with the rest of her life" tearjerkers are really only exploiting the gullible, who don't deserve to be abused the way the classic 419 victims do.