The Anon Coward parent article hits the nail pretty much on the head - there are people down there, even though they're mostly government-funded scientific expedition bases, as opposed to the native population, who care about fish, leopard seals, beer, and World Domination. The more practical answer is probably that they had an ISO Country Code (AQ) already assigned back when the Domain Name Gods decided that ccTLDs should use ISO country codes.
First of all, it made a lot of sense for them to expand the gTLD space using relatively boring names in the first batch - they'll only get to sell.INC once, and.LTD and.GMBH etc., and.XXX and.SEX give them two tries, and they had to do a lot of learning about the processes and the market. Much better to pick names nobody cares too much about, like.aero and.museum. Unfortunately, they waited way too long, and they'd have made a lot more money selling.INC in 1999 or 2000 than they can now that the boom is over (and if you don't think that money isn't as big an issue to ICANN as Intellectual Property trademark disputes are, you haven't been paying attention.)
But.BIZ is Great! 101% of the names in.BIZ belong to people that I have no interest in talking to, so any email that even contains a.BIZ URL can go straight to the spam-bucket. The only exception I've found is whois.biz, which is quasi-useful if I feel like hunting down a.biz spammer.
.MUSEUM is the only gTLD which seems to have done any design experiments in alternative ways to run a TLD besides simple hierarchies. It's not super-exciting, but it's at least something, and ICANN's policies about non-refundable $50,000 fee to apply for a gTLD mean that there won't be much experimentation possible.
Search engine dominance isn't an embedded infrastructure contest like railroad dominance - it's a popularity contest that can change in a heartbeat if something better comes along. Google became popular and crowded out its competition because it was a fundamentally friendlier engine, not in the sense of having syntactic sugar and flashy decorations like Hotwired, but in the sense of producing highly relevant results up front instead of mixing them randomly through the 50,000 matches for your search terms, indexing more of the web than most of the competitors, and being lightning fast as well, which it could do partly because its interface was lean and clean. Google as a business has some stickiness because of its popularity, which enabled it to raise enough cash in the market to hire the best and brightest to do new cool stuff, and they keep adding more cool stuff, and maybe some of that will add some business relationship stickiness that will keep other people around in addition to the popularity contest, but the fundamentals are still about having the quality it takes to maintain the popularity.
Perhaps Chirac can win part of the popularity contest in France by getting some academics and engineers to produce a service that's elegant, efficient, and French, something with the spirit of Eiffel as opposed to Inspector Clouseau or Derrida or De Gaulle or Sartre. Or perhaps he can pull off another Minitel - lightweight and pretty lame but good enough to get the job done given the lack of competition. And hopefully he can produce something that provides really good access to the information produced by the French government. But bureaucratic fiat isn't the way to produce popularity - you need a combination of luck, really really good technical skills, willingness to experiment, and a deep understanding of what your potential customers might want, and usually bureaucratic fiat produces things like bureaucrats and Fiats.
Oh, wait, you said "$500", not "$5.00":-) I did most of my game obsessing back when a 9600-baud modem made playing Nethack marginally faster compared to 2400, and Solitaire wastes just about as much time on WinXP with 2.4GHz CPU as it did on Win3.1.
It's nice to have enough graphics horsepower to watch DVDs, and whatever H.264 uses is theoretically useful as well, but mostly I just want a graphics card that can pump as many pixels as my monitor can handle at a speed that won't flicker, because I want good-looking text.
If you're asking for synchronous lines, they'd get confused. You sound like you're really talking about symmetric vs. asymmetric lines. It still does make sense for them to offer them, because the technology is cheaper, even though there are occasional people who want more upstream bandwidth - usually they'll get cheaper service by selling them a faster asymmetric service.
But what do you mean by "My ISP"? Do you mean your cable modem company, or do you mean some DSL company that you're using, or some DSL company your telco has a deal with?
Cable modem service really is highly asymmetric technology, and when they deploy faster services for businesses it's usually either DSL or a fat private-line access of some sort, which they'd typically deploy equipment for into heavily-business areas, not heavily-residential areas (if you're in a suburb, your local corner store probably can't get high-end business lines from the cableco either, just DSL.)
DSL services come in many flavors of symmetric and asymmetric, and most of the symmetric versions top out at 1.1 or 1.5 Mbps, and only if you're close enough to the telco office - I live somewhere between 11000 and 16000 feet from my telco, depending on which telco tech you believe, and I can only get SDSL up to 384kbps, while 768 fails - but the ADSL flavors generally get better distance, so I'm running on antique 1544/384 service until I get around to upgrading to 3 Mbps.
384 kbps is a *lot* of upstream bandwidth for a home user. Unless you're file-sharing, or trying to ship out Linux distributions, or running a business that ships lots of graphical data to prospective customers, you're not going to keep it very full. Business video-conferencing, for instance, typically uses 320-384kbps of video data, which expands to 400-450 after adding IP headers, but that's the kind of image you get from a $5-10,000 Polycom system; a $29 webcam is more likely to run 128kbps or less. VOIP runs as much as 80kbps if you're using uncompressed voice, so it can be nice to have 384 instead of just 128 upstream, but it's not critical. Blogging and posting on forums don't use any significant upstream bandwidth - you probably don't type faster than 100 WPM, which is about 100 bps, so maybe if you wrap a lot of web decoration around it you need 300 baud upstream. (If you're filesharing, then yes, you can soak up unlimited upstream bandwidth, especially if you're using BitTorrent, which is designed to use bandwidth aggressively.)
$50/hour is a bit cheap, but it's not out of line for outsourcing an entire department. It's certainly not the kind of rate you'd get for mid-level consultants you brought in for a specific project, much less security wizards for an emergency, but it's the kind of price you might charge to replace an IT staff that's mostly doing operations.
It's pretty tough to restore from known good backups, unless you make assumptions like "he didn't know the outsourcing/layoffs were coming until 2 months ago" and "hand-inspecting the backups is good enough". Might account for why they used 400 hours of grunt-work instead of security-wizard time.
The guy had significant access to their system, built in back doors, kept multiple access tokens around for a while after he was laid off, and used it to vandalize accounts. Who knows what else he left in the system, or whether he built backdoors into the backup systems, or left rootkits hanging around critical machines, or what other damage he might have done?
IBM apparently charged Adventis their standard "$50/hour outsourcer grunt" rate, not a "$2000/day medium-level consultant" rate or an "If you have to ask you can't afford it Security Wizard" rate. Not only did Adventis get off way light paying for the lower-priced consultants (though admittedly a lot of the work is scanning logfiles, if the logfiles can be trusted), but either the system was designed to really effectively limit the scope that he had access to, which is a dodgy assertion if he had anything to do with designing it, or else they should have brought in much bigger guns to find out what he might have tampered with. (Of course, they should have also had backups for the critical information that they could pop up quickly, and the probably did, so hopefully most of the work was done after they'd restored access to the other sysadmin, but can you trust the backups?) Sometimes destructive people are just opportunistically trashing whatever's nearby, and maybe they decided that that was all he did, but if he'd been seriously trying to sabotage them he could have caused a lot more damage.
If you RTFA, his former employer hired IBM to administer computers for them, and dumped some of their direct employees including him. IBM is apparently billing $50/hour for labor, and recorded 407 hours of labor that was charged to this project and billed to Adventis, and that's the kind of project work that's part of the standard billing arrangement for this sort of computer outsourcing. IBM certainly won't report this as a loss - it was billable work charged to their customer, though for Adventis this is a loss that might show up on a balance sheet if it's only rounded to the nearest thousand and not the nearest million. He's also getting off way light on the costs - IBM was apparently charging this as a typical US outsourcing "Grunts by the hour" price of $50, not a $2000/day "mid-level consultant" rate or a $5-10K/day "security wizard" rate - while much of the work was crunching through log files, doing a thorough cleanup job means looking for deep penetration of backups and access systems. He could have easily been hit for a couple hundred thousand.
The guy deliberately kept passwords and access devices for a system he'd been responsible for, and deliberately trashed parts of the system and deleted accounts for other administrators, and he deserves what happens to him. This isn't like Mitnick giving away information, or even crackers using the victim's machine as a launching pad for zombies - it's pure premeditated vandalism. The concept of a "protected computer" in Federal laws may be dodgy, but he did a lot more real and potential damage than stealing a company car, a crime for which nobody would be bothered by him getting a few months in jail.
If anybody's ripping anybody off here, it's his lawyers taking this to a Federal Appeals Court when the guy's obviously getting off light, and you know his lawyers are charging him a lot more than $50/hour and billing a lot more hours if they're getting to that level of the courts. They should have told him to do a plea-bargain and helped him get one that avoids jail time, but maybe the initial judge wouldn't go for it and he thought it was worth the money to try to get bounced to a state court.
I'm happy to accept pointers to better documentation. But in fact the eTree distributions do have separate files for each track, and generally have a text file that describes each track as well. The problem is finding tools that can use them (oh, and they have to run under Windows - the support does look better under Linux.)
That's chemical cracking, not nuclear cracking
on
Return to the Moon
·
· Score: 1
Sure, it's theoretically possible to find various sets of nuclear fission that'll let you get hydrogen and maybe oxygen out of other nuclei, though it's not clear it'd ever be cost-effective (especially if there isn't a lot of convenient uranium lying around on the moon.)
But the much more reasonable industrial process that cracks stuff into H, O, and leftovers is chemical cracking that uses the nuclear reactor as a heat source, so you can do things that are endothermic. On the moon, that might let you crack oxygen out of various kinds of rock, but if there's no hydrogen in the rocks, then you're not going to get hydrogen. If there's enough nuclear fuel around, it might still be useful, but you've also got the sun as a heat source - it may make more sense to concentrate it using big mirrors, and use the leftover Si/Al/etc. from teh rocks to make more mirrors.
Funny you should mention that. I've got some music in my iTunes downloaded from the store, some ripped from CDs I own, and some that I downloaded from eTree in FLAC or SHN lossless-compression formats (legal jam-band concert downloads.) The tools for converting lossless formats into MP3 or AAC seem to have real problems keeping track of information about the music - they know it's "gdead 1995-06-27 - track 4", but don't seem to have a way to import the song names that's better than doing cut&paste. It's quite annoying.
I just received an iPod (because my employer prefers to reward performance with consumer electronics as opposed to actual cash, but whatever:-) Once you get past the outer packaging, the iPod's wrapped in a plastic layer that says "Don't Steal Music" in several languages, so I've been warned. Now if I could just get the battery to charge correctly:-)
Email itself is very cheap, and uses up a trivial fraction of your bandwidth and CPU unless you're running a business-sized email server, even if you don't filter out all of the spam. (It can use more than a trivial fraction if you're running really really aggressive filters, but that's rare.) Advertising banners soak up a lot more of your bandwidth and horsepower, as does web browsing.
The problem with spam is not that it's costing you money - it's that it's wasting your time. Your email provider is probably spending a lot of resources trashing spam, but except for the largest providers, it's personnel time that's costing them money, not equipment, and they're doing it because they want to keep your business.
That doesn't mean that spammers aren't scum, because of course they are. But these laws don't just affect spammers - they also attack *anybody* who's emailing politically-incorrect-by-Utah-standards information that might be received by a minor in Utah, forcing them to pay the state's subcontractor an unreasonable price to listwash their entire mailing lists unless they have other ways to prove that none of their subscribers are Utah teenagers - and they're not able to defend themselves merely by having a 'Check this box to indicate that you're not a Utah teenager', because teenagers are presumed to lie about that sort of thing, especially if they want pr0n or cigarettes. One of the affected parties was a wine seller who doesn't even ship products to Utah, but they're forced to deal with the law. And after all, it would be horrible if teenagers were exposed to *gambling* by email, as opposed to having to go to a website that was advertised on TV, or to learn poker in Boy Scouts or sports betting at the high-school lunchtable like I did (our group of geeks then used the school's computer lab to adapt the football coach's winner-prediction program to predict winners of professional games as opposed to strategies for high school games:-)
If anybody in the world sends politically-incorrect-for-Utah-minors information to a kid in Utah, this law says they're guilty, whether the kid requested it or not, and whether the kid lied about his age or location on a form or not. One of the alcoholic-beverage merchants was especially concerned about this, because while they don't even _sell_ products in Utah, they have to spend 5 cents per subscriber for anybody they don't know for certain not to be a kid in Utah, and another 7 cents for Michigan, to avoid the risk of fighting a criminal charge and heavy fines. And the people who are hardest hit are the legitimate US businesses - spammers can scam people just as easily from offshore, with at most the cost of an extra $100 corporate registration and a Chinese website.
Yes, it'd be nice if spammers couldn't afford to send email to Utahns, or to Californians like me. It'd be extremely annoying if California wineries couldn't send information to people who ask for it, or small beer brewers anywhere, or if private websites that provide free correct information about politically incorrect drugs couldn't provide it (or even if lying scum like the Partnership for a Drug-Free America couldn't do that).
Much of the information they're censoring is covered by the Commerce Clause of the US constitution as interstate commerce, which states aren't allowed to regulate. Alcohol's a special leftover from Prohibition, and the Feds have their own laws about gambling which several scummy Congresscritters are in favor of, and while I wouldn't mind if I stopped getting told how to collect my winnings in the Nigerian Lottery, that's also out of their jurisdictions.
Utah and Michigan each have some service that they contract to who gets an encrypted list of the block-list, and who bulk mailers can send their mailing lists to to be validated, for $5/1000 in one state and $7/1000 in the other. That means that if you don't know that an email address isn't in Utah or Michigan, and you're sending politically incorrect material, you need to spend 12 cents to validate that it's not a child in either state - even if you were trying to run a perfectly legitimate subscription-based email system for people who told you they wanted to receive your email or buy your products and you were trying to send them the confirmation email to make sure it was really them.
A hash list could work in a more convenient manner, though it's susceptible to dictionary attacks. For instance, you can figure out the domain names, especially because potential customers for spam products usually use big email services instead of small clueful ones, so you can just start hashing likely usernames at each of the top 100 mail systems. Finding specific addresses is fast, but as long as you're just trying to find lots of addresses, and not trying to find _all_ of them, it's also easy to do that. But it is a good start.
The fun part for the people running the list is to seed it with lots of trap addresses, and see who bites. But you don't actually have to be running the list to seed it - anybody who's jurisdictionally eligible to put their email addresses on the list can include bait as well as addresses they actually use.
Dictionary attacks used to be very common - just hit a few hundred million likely names at whatever domain you're targeting. My company used to provide secondary-MX service for one of my customers, back around the time that spammers were discovering that almost nobody ran spam-prevention on secondary MXs and almost everybody whitelisted their secondary MX server, and we had a month of every-weekend attacks of tens of millions of messages, which would overrun the undersized spam-filter machine that was front-ending their (old) Exchange server and didn't have access to the list of valid email addresses, so it'd have to accept each one and search it for keywords instead of blocking it by IP or RCPT TO: header. We both decided that the $50/month add-on feature shouldn't be allowed to jeopardize the main services we were providing for them, and killed off the secondary MX. (We've got bigger mail-handling systems now; that was back when secondary MX service was useful and nearly everybody had it and spam was an appalling 20-30% of total email traffic.)
My primary email address has been wandering around the net in mailing list archives and Usenet for over a decade, and even though I'd never use it to sign up for stuff I don't want, it's been extensively hit by harvesters. One email address from a cryptographic mail server I stopped running in ~1997 still gets spam, even though I never even sent email from it except as a response to email it received - it was mentioned on a web page about the project, so harvesters got it, and they kept selling the address to each other for a couple of years after I took it off the web page. (That was back when lots of the spam was for Millionsss of Fresssh New Email Addressess, so advertising it to an old stale address seemed appropriate.)
Besides, if you want to give an email address to somebody you expect will spam you, that used to be what Hotmail was for, and still is what Yahoo and other free email services are good for, or for other kinds of services, bugmenot/dodgeit/mailinator/etc. do even more disposable address types. It's also useful to do tagged email addresses (username+tag@example.com, or tag@username.example.com.) The latter format has the advantage that it can increase the numbers of addresses a spammer has to try by several orders of magnitude, depending on whether the spammer can find the subdomain names easily, though it does occasionally mean that your subdomain will get dictionary-spammed.
It's historically unclear where the Ladies' Sewing Circle and Terrorist Society meme and logo originated - seems to be the 1970s or before, and you're probably not allowed to wear the T-Shirt in airports any more. But you can apparently carry knitting needles on airplanes again, or at least somebody sitting next to me last week had done so. Maybe you just need to use plastic or wooden ones?
Hey, the spammers can get PCs using [fill in your favorite chip maker] CPUs, and they'll stay toasty warm. I used to heat one of my labs with a couple of Sun-4s, and another one with a VAX 11/780. (Actually, I'd happily recommend that they get a VAX to do their spamming from - it's much more efficient, though a lot slower, and some of them will get electrocuted trying to install 208-volt 3-phase power....)
As far as I can tell the technical details from the WSJ article, it's not a simple list - there's some contractor that receives the data in encrypted form and manages the list, and if you want to validate against the list you need to pay them about $5 per 1000 addresses, and you get feedback about whether they are or are not on the Utah list, and there's a similar deal for $7/1000 for the Michigan list. So if you're selling politically incorrect material, whether it's porn and gambling spam or whether it's a subscription-only wine newsletter, and you don't want to risk becoming a criminal, you have to pay to validate *all* the names on your lists, just in case some kid from Utah or Michigan might be on them.
While the law does attack spammers, it also attacks confirmed opt-in subscriptions - if you RTFA, you'll see that the beverage sellers were complaining that they have to pay Utah 5 cents to certify every name on their list, even though they don't sell products in Utah, just in case some kid from Utah tries to subscribe to their newsletter.
If you read TFA, you'll see that one of the alcohol-sellers said that the law forced them to pay two separate states to verify their entire subscriber list because some kid from Utah _might_ request to subscribe, and it would be illegal for them to accept the subscription. And the combined price for the two states is 12 cents per name, which is really annoying for a free newsletter.
It's not just about spam - it's about all kinds of speech, and about the technical competence of the lawmakers, who don't understand the implications of the laws they're writing.
Apparently dogs are much much harder to clone than sheep or cats - there are just lots of weird things that go on which make it difficult, so if that part wasn't fraud, then he's accomplished something useful before trashing his reputation and prospects of future work.
There was a classic article from ~1984 about "Why did I get a Cairn Terrier instead of a Macintosh?" A Cairn Terrier is a little dog like Toto, and the article went on to compare price, memory capacity, voice recognition, upgradability, friendliness, etc. The little dog won, but Macs were pretty new back then.
Slashdot Mirror
For a non-refundable $50K donation to ICANN, you can get them to evaluate your business plan for .OGR - you're not allowed to troll them for free...
The Anon Coward parent article hits the nail pretty much on the head - there are people down there, even though they're mostly government-funded scientific expedition bases, as opposed to the native population, who care about fish, leopard seals, beer, and World Domination. The more practical answer is probably that they had an ISO Country Code (AQ) already assigned back when the Domain Name Gods decided that ccTLDs should use ISO country codes.
But .BIZ is Great! 101% of the names in .BIZ belong to people that I have no interest in talking to, so any email that even contains a .BIZ URL can go straight to the spam-bucket. The only exception I've found is whois.biz, which is quasi-useful if I feel like hunting down a .biz spammer.
.MUSEUM is the only gTLD which seems to have done any design experiments in alternative ways to run a TLD besides simple hierarchies. It's not super-exciting, but it's at least something, and ICANN's policies about non-refundable $50,000 fee to apply for a gTLD mean that there won't be much experimentation possible.
Perhaps Chirac can win part of the popularity contest in France by getting some academics and engineers to produce a service that's elegant, efficient, and French, something with the spirit of Eiffel as opposed to Inspector Clouseau or Derrida or De Gaulle or Sartre. Or perhaps he can pull off another Minitel - lightweight and pretty lame but good enough to get the job done given the lack of competition. And hopefully he can produce something that provides really good access to the information produced by the French government. But bureaucratic fiat isn't the way to produce popularity - you need a combination of luck, really really good technical skills, willingness to experiment, and a deep understanding of what your potential customers might want, and usually bureaucratic fiat produces things like bureaucrats and Fiats.
I did most of my game obsessing back when a 9600-baud modem made playing Nethack marginally faster compared to 2400, and Solitaire wastes just about as much time on WinXP with 2.4GHz CPU as it did on Win3.1.
It's nice to have enough graphics horsepower to watch DVDs, and whatever H.264 uses is theoretically useful as well, but mostly I just want a graphics card that can pump as many pixels as my monitor can handle at a speed that won't flicker, because I want good-looking text.
But what do you mean by "My ISP"? Do you mean your cable modem company, or do you mean some DSL company that you're using, or some DSL company your telco has a deal with?
Cable modem service really is highly asymmetric technology, and when they deploy faster services for businesses it's usually either DSL or a fat private-line access of some sort, which they'd typically deploy equipment for into heavily-business areas, not heavily-residential areas (if you're in a suburb, your local corner store probably can't get high-end business lines from the cableco either, just DSL.)
DSL services come in many flavors of symmetric and asymmetric, and most of the symmetric versions top out at 1.1 or 1.5 Mbps, and only if you're close enough to the telco office - I live somewhere between 11000 and 16000 feet from my telco, depending on which telco tech you believe, and I can only get SDSL up to 384kbps, while 768 fails - but the ADSL flavors generally get better distance, so I'm running on antique 1544/384 service until I get around to upgrading to 3 Mbps.
384 kbps is a *lot* of upstream bandwidth for a home user. Unless you're file-sharing, or trying to ship out Linux distributions, or running a business that ships lots of graphical data to prospective customers, you're not going to keep it very full. Business video-conferencing, for instance, typically uses 320-384kbps of video data, which expands to 400-450 after adding IP headers, but that's the kind of image you get from a $5-10,000 Polycom system; a $29 webcam is more likely to run 128kbps or less. VOIP runs as much as 80kbps if you're using uncompressed voice, so it can be nice to have 384 instead of just 128 upstream, but it's not critical. Blogging and posting on forums don't use any significant upstream bandwidth - you probably don't type faster than 100 WPM, which is about 100 bps, so maybe if you wrap a lot of web decoration around it you need 300 baud upstream. (If you're filesharing, then yes, you can soak up unlimited upstream bandwidth, especially if you're using BitTorrent, which is designed to use bandwidth aggressively.)
$50/hour is a bit cheap, but it's not out of line for outsourcing an entire department. It's certainly not the kind of rate you'd get for mid-level consultants you brought in for a specific project, much less security wizards for an emergency, but it's the kind of price you might charge to replace an IT staff that's mostly doing operations.
It's pretty tough to restore from known good backups, unless you make assumptions like "he didn't know the outsourcing/layoffs were coming until 2 months ago" and "hand-inspecting the backups is good enough". Might account for why they used 400 hours of grunt-work instead of security-wizard time.
IBM apparently charged Adventis their standard "$50/hour outsourcer grunt" rate, not a "$2000/day medium-level consultant" rate or an "If you have to ask you can't afford it Security Wizard" rate. Not only did Adventis get off way light paying for the lower-priced consultants (though admittedly a lot of the work is scanning logfiles, if the logfiles can be trusted), but either the system was designed to really effectively limit the scope that he had access to, which is a dodgy assertion if he had anything to do with designing it, or else they should have brought in much bigger guns to find out what he might have tampered with. (Of course, they should have also had backups for the critical information that they could pop up quickly, and the probably did, so hopefully most of the work was done after they'd restored access to the other sysadmin, but can you trust the backups?) Sometimes destructive people are just opportunistically trashing whatever's nearby, and maybe they decided that that was all he did, but if he'd been seriously trying to sabotage them he could have caused a lot more damage.
The guy deliberately kept passwords and access devices for a system he'd been responsible for, and deliberately trashed parts of the system and deleted accounts for other administrators, and he deserves what happens to him. This isn't like Mitnick giving away information, or even crackers using the victim's machine as a launching pad for zombies - it's pure premeditated vandalism. The concept of a "protected computer" in Federal laws may be dodgy, but he did a lot more real and potential damage than stealing a company car, a crime for which nobody would be bothered by him getting a few months in jail.
If anybody's ripping anybody off here, it's his lawyers taking this to a Federal Appeals Court when the guy's obviously getting off light, and you know his lawyers are charging him a lot more than $50/hour and billing a lot more hours if they're getting to that level of the courts. They should have told him to do a plea-bargain and helped him get one that avoids jail time, but maybe the initial judge wouldn't go for it and he thought it was worth the money to try to get bounced to a state court.
I'm happy to accept pointers to better documentation. But in fact the eTree distributions do have separate files for each track, and generally have a text file that describes each track as well. The problem is finding tools that can use them (oh, and they have to run under Windows - the support does look better under Linux.)
But the much more reasonable industrial process that cracks stuff into H, O, and leftovers is chemical cracking that uses the nuclear reactor as a heat source, so you can do things that are endothermic. On the moon, that might let you crack oxygen out of various kinds of rock, but if there's no hydrogen in the rocks, then you're not going to get hydrogen. If there's enough nuclear fuel around, it might still be useful, but you've also got the sun as a heat source - it may make more sense to concentrate it using big mirrors, and use the leftover Si/Al/etc. from teh rocks to make more mirrors.
Funny you should mention that. I've got some music in my iTunes downloaded from the store, some ripped from CDs I own, and some that I downloaded from eTree in FLAC or SHN lossless-compression formats (legal jam-band concert downloads.) The tools for converting lossless formats into MP3 or AAC seem to have real problems keeping track of information about the music - they know it's "gdead 1995-06-27 - track 4", but don't seem to have a way to import the song names that's better than doing cut&paste. It's quite annoying.
also, this might be a first post...
The problem with spam is not that it's costing you money - it's that it's wasting your time. Your email provider is probably spending a lot of resources trashing spam, but except for the largest providers, it's personnel time that's costing them money, not equipment, and they're doing it because they want to keep your business.
That doesn't mean that spammers aren't scum, because of course they are. But these laws don't just affect spammers - they also attack *anybody* who's emailing politically-incorrect-by-Utah-standards information that might be received by a minor in Utah, forcing them to pay the state's subcontractor an unreasonable price to listwash their entire mailing lists unless they have other ways to prove that none of their subscribers are Utah teenagers - and they're not able to defend themselves merely by having a 'Check this box to indicate that you're not a Utah teenager', because teenagers are presumed to lie about that sort of thing, especially if they want pr0n or cigarettes. One of the affected parties was a wine seller who doesn't even ship products to Utah, but they're forced to deal with the law. And after all, it would be horrible if teenagers were exposed to *gambling* by email, as opposed to having to go to a website that was advertised on TV, or to learn poker in Boy Scouts or sports betting at the high-school lunchtable like I did (our group of geeks then used the school's computer lab to adapt the football coach's winner-prediction program to predict winners of professional games as opposed to strategies for high school games :-)
Yes, it'd be nice if spammers couldn't afford to send email to Utahns, or to Californians like me. It'd be extremely annoying if California wineries couldn't send information to people who ask for it, or small beer brewers anywhere, or if private websites that provide free correct information about politically incorrect drugs couldn't provide it (or even if lying scum like the Partnership for a Drug-Free America couldn't do that).
Much of the information they're censoring is covered by the Commerce Clause of the US constitution as interstate commerce, which states aren't allowed to regulate. Alcohol's a special leftover from Prohibition, and the Feds have their own laws about gambling which several scummy Congresscritters are in favor of, and while I wouldn't mind if I stopped getting told how to collect my winnings in the Nigerian Lottery, that's also out of their jurisdictions.
A hash list could work in a more convenient manner, though it's susceptible to dictionary attacks. For instance, you can figure out the domain names, especially because potential customers for spam products usually use big email services instead of small clueful ones, so you can just start hashing likely usernames at each of the top 100 mail systems. Finding specific addresses is fast, but as long as you're just trying to find lots of addresses, and not trying to find _all_ of them, it's also easy to do that. But it is a good start.
The fun part for the people running the list is to seed it with lots of trap addresses, and see who bites. But you don't actually have to be running the list to seed it - anybody who's jurisdictionally eligible to put their email addresses on the list can include bait as well as addresses they actually use.
My primary email address has been wandering around the net in mailing list archives and Usenet for over a decade, and even though I'd never use it to sign up for stuff I don't want, it's been extensively hit by harvesters. One email address from a cryptographic mail server I stopped running in ~1997 still gets spam, even though I never even sent email from it except as a response to email it received - it was mentioned on a web page about the project, so harvesters got it, and they kept selling the address to each other for a couple of years after I took it off the web page. (That was back when lots of the spam was for Millionsss of Fresssh New Email Addressess, so advertising it to an old stale address seemed appropriate.)
Besides, if you want to give an email address to somebody you expect will spam you, that used to be what Hotmail was for, and still is what Yahoo and other free email services are good for, or for other kinds of services, bugmenot/dodgeit/mailinator/etc. do even more disposable address types. It's also useful to do tagged email addresses (username+tag@example.com, or tag@username.example.com.) The latter format has the advantage that it can increase the numbers of addresses a spammer has to try by several orders of magnitude, depending on whether the spammer can find the subdomain names easily, though it does occasionally mean that your subdomain will get dictionary-spammed.
It's historically unclear where the Ladies' Sewing Circle and Terrorist Society meme and logo originated - seems to be the 1970s or before, and you're probably not allowed to wear the T-Shirt in airports any more. But you can apparently carry knitting needles on airplanes again, or at least somebody sitting next to me last week had done so. Maybe you just need to use plastic or wooden ones?
Hey, the spammers can get PCs using [fill in your favorite chip maker] CPUs, and they'll stay toasty warm. I used to heat one of my labs with a couple of Sun-4s, and another one with a VAX 11/780. (Actually, I'd happily recommend that they get a VAX to do their spamming from - it's much more efficient, though a lot slower, and some of them will get electrocuted trying to install 208-volt 3-phase power....)
As far as I can tell the technical details from the WSJ article, it's not a simple list - there's some contractor that receives the data in encrypted form and manages the list, and if you want to validate against the list you need to pay them about $5 per 1000 addresses, and you get feedback about whether they are or are not on the Utah list, and there's a similar deal for $7/1000 for the Michigan list. So if you're selling politically incorrect material, whether it's porn and gambling spam or whether it's a subscription-only wine newsletter, and you don't want to risk becoming a criminal, you have to pay to validate *all* the names on your lists, just in case some kid from Utah or Michigan might be on them.
While the law does attack spammers, it also attacks confirmed opt-in subscriptions - if you RTFA, you'll see that the beverage sellers were complaining that they have to pay Utah 5 cents to certify every name on their list, even though they don't sell products in Utah, just in case some kid from Utah tries to subscribe to their newsletter.
It's not just about spam - it's about all kinds of speech, and about the technical competence of the lawmakers, who don't understand the implications of the laws they're writing.
Apparently dogs are much much harder to clone than sheep or cats - there are just lots of weird things that go on which make it difficult, so if that part wasn't fraud, then he's accomplished something useful before trashing his reputation and prospects of future work.
There was a classic article from ~1984 about "Why did I get a Cairn Terrier instead of a Macintosh?" A Cairn Terrier is a little dog like Toto, and the article went on to compare price, memory capacity, voice recognition, upgradability, friendliness, etc. The little dog won, but Macs were pretty new back then.