I don't get it--since when do we tax assets? We don't do that here in the US, and I didn't think that Europe did, either. Keeping 2 years worth of widgets on hand isn't expensive because of taxes, it's expensive because of the storage space it takes up.
But on top of that, companies already DO pay for the government's "protection": they pay corporate income taxes, sales taxes, and more.
And it's kind of silly, from a tax standpoint, to let the individual companies determine the value of their own intellectual property. Once an asset is "on the books", a company can value it however they want, which may result in some absurd claimed valuations depending on how they do their accounting. For instance, the company that owns the Empire State Building in NYC lists the asset on its general ledger as being worth exactly $1. No, NOT $1 million, but one single dollar. They're perfectly allowed to do that, and there is a sensible reason for it as far as how they initially dealt with the costs of building the place.
So if you say "pay %1 maintenance for intellectual property", the big media companies will just adjust their booked asset values downward. You could always pass a law preventing them from doing that, but such a law would have lots of horrible unintended consequences, because the ability to book arbitrary asset values is a necessary part of accounting.
Um, no. MS could easily sell you Windows without asking you to agree with a EULA/license. But they do.
They have the right to forbid others from using their software, because they own the copyright. That's what allows them to demand payment: "If you don't pay us $300, you can't use Windows." They also demand that you agree to the EULA, in order to be allowed to install the software on your machine.
Now, the legality of some of these implementations (click-thru licenses, pre-installed licenses) has been questioned, but the basics are known to be true:
1) If I own the copyright on a work, I can forbid others from using the it.
2) In exchange for being allowed to use my works, I can demand arbitrary concessions from other parties.
No, because unless you agree to a license/contract when you buy it, they can't enforce any terms on you, regardless of the fact that they own a copyright on the software that runs on the camera. You're not allowed to copy the copyrighted software on the camera without their permission, but that's the limit of what they can claim without a license agreement.
My coffee maker has software--it's a real flash job. I'm not allowed to copy that software, but they can't claim that any code I write while drinking the coffee made with it is a "derived work"... Maybe that's a cheeky example.
MS owns the copyright on Windows. They agree to let you use their software IF you agree to their EULA/license. If they just sell you the software without making you agree to the license, they have no power over you other than to prevent you from making unauthorized copies.
A good comment, but you didn't quite get what I'm asking in the OP.
What the DMCA does prevent is anybody (like Adobe) from "traficking" in software which decrypts the data. It doesn't matter who actually owns the copyright...even if the photographer releases everything to the Public Domain. It's the fact that there could exist any copyrighted photo which Adobe's software could be used to decrypt by anybody else who's not the copyright holder.
THAT'S what the question is: does the DMCA make encryption tools unlawful that are used to decrypt data to which the decrypting party own the copyright? That's not clear at all, to me. Adobe THINKS that it may be the case, and so they've decided to scoot around the issue... But nobody's actually established to me, yet, that this is really the case!
Look at it this way: it's NOT illegal to decrypt any encrypted information. You can write a WEP-cracking tool, or a Kerberos password brute-forcer, or Jack the Ripper, and you don't run afoul of the DMCA. The DMCA only comes into effect when you're talking about breaking encryption to violate copyright.
Adobe could well be acting over-carefully, here, just to avoid the issue entirely.
Sorry, Charlie. You missed this boat. Too bad you didn't RTFA.
in the very quote you provided, Nikon claims to own the copyright on that information
The guy who wrote the PhotoshopNews article made the statement that "...Nikon apparantly thinks they own the information..." (emphasis mine). As far as any of the article says, or any of the links reveal, Nikon hasn't said ANYTHING about this case. Adobe's employees said something, but not Nikon.
The auther is assuming from the Adobe people's statements that they think that Nikon feels a certain way. But Nikon never actually said that. And if Nikon isn't claiming a copyright on the data in the file, they could still encrypt it in order to make 3rd-party software less useful. Nothing stops them from being bastards.
And now that I think about it, why IS the white balance information any different from the image? The image is just bits, which make a recognizable image when interpreted a certain way by image software. The copyrighted form of the image IS the bits that describe it. Some of those bits are encoded in the white balance information--they help describe the image. Just because most cameras don't report the white balance data doesn't mean that it isn't "part of the image."
They likely retain their copyright over the "NEF" part even if it is incorporated into the subsequent photograph.
That doesn't make any sense, though. You don't have a EULA when you buy a camera, because they're not licensing the camera for you to use. They're SELLING you the camera, which you then own. There's no copyright over the photographs that Nikon ever holds that would allow them to "retain" some portion of it.
What you're describing is like a typewriter manufacturer claiming copyright on anything typed on their equipment, or a refrigerator maker claiming copyright on the pictures that my kids draw that I hang on the front of it. Since there's no agreement between me and the manufacturer that I'm assigning copyrights to them as a condition of using their product, they can't possibly own copyrights on the work.
I'm not saying that Nikon COULDN'T have such an agreement with the camera (maybe, maybe not), just that I'm pretty sure no such agreement exists when you buy one.
So it's a way for them to lower the retail price of the camera, similar to the way rebates work.
That doesn't make ANY sense. This is a $5,000+ camera--an extra $100 on the retail price isn't going to have shit to do with anyone's decision to buy it. When you're buying high-end stuff like this, you're probably not making a decision based solely on comparative prices, and you're also probably not THAT sensitive to price that you're care about a 2% difference. When people buy Bentley and Rolls-Royce cars, they don't care that the CD player costs $500 more than an after-market shop would charge, because they're dropping $100K on the car, anyway.
More likely, Nikon wanted to add the raw shooting mode to the camera, requiring white balance data in the file format, but wasn't entirely sure that software on the market (like Photoshop) would properly support the proprietary stuff when the camera hit the market. So they spent the cash to write their own software in-house. Then, somebody says "why are we giving the software away for free?" and they decide to charge for it, which justifies a decision to encrypt the data in the file to force the business model on the consumer.
Which looks like a dumb decision from here, but that's probably because different people made different parts of the decision, probably without stopping to consider the bigger picture: no professional photog is going to want to use their shitty proprietary software, if they can help it.
Not exactly. Assuming that Nikon actually owns the rights to the algorithm that they're using for the encryption (they might not, instead using someone else's or a public domain algorithm), licensing it doesn't get Adobe anywhere. The DMCA doesn't say you're not allowed to use an algorithm, it says you're not allowed to actually decrypt the data.
I think you meant that Adobe should get an agreement with Nikon that they're allowed to build and sell the decryption software, and that Adobe's customers are allowed to use it on their photos. That's not a licensing agreement, exactly... I don't know what you'd call it.
(I personally think that would be a bogus interpretation of the DMCA, since I think the copyrighted information inside the NEF file belongs to the photographer, not Nikon. But Nikon apparently thinks they own the information inside the NEF).
This is a little strange, isn't it? If a photographer takes a picture, it's pretty clear that the photog owns the copyright to that photo. Nikon couldn't possibly claim any rights on photos taken with their camera, least of all because it would make it impossible for professionals to use that equipment. And with $5K cameras, you're really only looking at the professional market.
So if the white balance information (the encrypted stuff) is a part of the photograph, the photographer owns the copyright on that data, too, right? That seems pretty straightforward, but I could be wrong...
Can the DMCA be applied to prevent you from decrypting something that you own the copyright on? This isn't even like owning a DVD and wanting to decrypt the data, because in that case the movie company owns the copyright.
If the DMCA can be applied that way, that's some fucked-up shit. It's just absurd.
RMS may be a little freaky... Okay, he's a LOT freaky. I think that most people are put off more, though, by the near-religious aspects of his approach to free software.
I'm a big free software fan, but I frame it in highly practical terms. Copyright is a monopoly, by definition--to some extent, it's a justifiable monopoly, because of the benefits of guaranteeing an economic return to the parties that create and fund its development. It helps to economically rationalize the costs of software development, because the work of creating it can be rewarded in a way that would otherwise be difficult.
But the downside of the monopoly is pretty big, too--proprietary lock in is the big one, in my mind, but there's more. If you're reliant on closed-source software, you usually have to take it as-it, bugs and missing features and bloat, and all. OSS tends to encourage a lot more flexibility, interoperability, and choice in the way things work. The business benefits of these freedoms are huge, no question.
The great thing about OSS is that it still allows revenue streams for software creators and funders. The creators of MySQL get a lot of consulting business based on the fact that they know the software better than anyone else. People who make big Linux kernel contributions can find jobs easily with companies that need Linux expertise, because they have a proven qualification in that area that means more than a million certifications. So the people that contribute have a payoff.
And GPL-style licenses are more effective at preserving these revenue opportunities than BSD-type licenses, because they tend to prevent proprietary code-forking, which keeps projects under the control of the original creators more easily.
This is the great realization of OSS that can be couched in purely economic terms: all the benefits of copyrightable works, but minimizing the negatives of the monopoly that comes with it.
I don't know how much RMS agrees with this particular sentiment, but then again, I don't really care.
Re:EMR from high tension power lines?
on
Quantum Wires
·
· Score: 4, Informative
I assume you're talking about the different effects of resistance on AC and DC currents: as electricity travels through a conventional conductor, the resistance of the conductor gives up some of the electrical power as heat (as Ohm's law describes). That's why we use high-voltage AC to distribute electric power, and even higher-voltage AC to transmit power over long distances--by transmitting at high AC voltages, you don't lose quite as much power as you otherwise would.
So if you could replace vast swaths of conventional copper electric transmission and distribution lines with superconductors, you could theoretically switch to DC power in these applications, which would have some interesting effects on the rest of the electrical distribution system.
Strict DC voltage on the power lines would virtually eliminate the EM radiation. You would still get some EM when you turned things on or off, or if the amount of power the line carries changed at all, but there would be a HELL of a lot less.
Lower voltages could be used, which would be safer (less chance of electrocuting people), and connectors (plugs, receptacles) designed with lowe voltages in mind would be cheaper to produce and certify.
Also, many devices in the home (especially computer equipment, or anything with circuit logic in it) need to convert the 110V AC current into much lower-voltage DC (2-5V DC, usually) to operate chip logic. This in generally an inefficient process, with a lot of power given up in the transformers and inverters to heat. Granted, you'd have to redesign all the home devices that currently use AC power directly (mostly lights and appliances) to run on DC, it could be done.
Really, the only problem would be the massive costs of switching over from one standard to another. All of those applicances and such would become useless on the new standard, which means everybody has to go out and buy new stuff. If you tried to switch the distribution over to DC in one go, I can see a lot of people having a lot of problems with it. And it wouldn't be practical to change the distribution bit-by-bit, either.
If you just wanted to change the transmission side, and leave the consumer out of it entirely, you'd have to replace a lot of power generation infrastructure. This could be done more slowly, I'd imagine, but it would still be expensive.
But then again, there's nothing that prevents you from continuing to run AC current on superconducting wires. That's probably what will happen, because it's the cheapest option.
I don't see anyone caring too much about interference from power lines in the 60Hz frequency band, anyway--not like we use those frequencies for anything.
Have you ever heard of a little thing called the "International Space Station"? Yeah, exactly.
If you leave Earth's protective envelope, you do become subject to larger exposures of radiation, but the danger can be minimized with shielding. People can take a decent amount of radiation exposure before they show any health effects, even the long-term cancer risks.
Basically, you just have to build sheltered structures for the inhabitants to block the radiation. It can be a choice of construction materials, or you can just bury the structures and pile regolith on top--the old bomb shelter solution. All it takes is a bulldozer and some internal supports, and you've got a pretty nice shelter.
Spacesuits aren't very good radiation protection, so they'll probably impose daily, monthly, and mission-specific limits on outdoor activities for inhabitants. If they also keep alert to solar activity, they can just head indoors when the worst stuff is coming.
How can you keep alert to incoming radiation, you ask? The ionizing, harmful parts of solar radiation are mostly charged particles, which travel slower than the speed of light. Big emissions of charged particles happen in conjunction with particular types of electromagnetic radition, which DOES travel at the speed of light (duh). So we look for the EM radiation that signals a coming charged particle storm, and tell the moonies to get indoors quick.
Not perfectly safe, but come on: they're on the fucking moon.
The statistics I barely remember from a year ago don't hold much weight, I'll grant you -- but the point is that we don't have to trade anecdotes, we can look at actual case studies. My money is on free wifi translating to better business, and for-pay wifi adding expenses that significantly limit its profitability. In that context, the GP is correct -- wifi providers would do well to get their minds right.
Seems like a lot of effort, given that I've gotSo let me start off by saying "There is a very large chance that you may be correct".
The problem with the OP's post wasn't about whether WiFi providers/locales have chosen the best business model: it's whether one person's assessment that a service costs too much for too little is a credible proof of whether the pricing model is good. That's an absurd conclusion. the fact that a Rolls-Royce limo is too expensive for me to consider buying one has NOTHING to do with the price of tea in China.
I use a LOT of free hotspots--tiny, 2-person startup company, no real office at the moment, so we work out of coffee shops and bars a lot.
Commercial hotspots have one thing that free nodes can't beat: quality of service. In terms of distance to the AP and interference from intervening objects, you'll usually get a much better signal from the commercial AP. This is exacerbated in a built-up city like NYC, because the buildings severely cut down on wireless propagation between blocks--normally, you'll only get strong enough signals to connect to something on the same block, with a rough line-of-sight to the AP.
But I can see that being a non-issue in suburban areas, or the cities out West where the buildings are smaller, fewer, shorter, and lighter of construction. I've noticed in other areas that the propagation issues are balanced by the fact that shit is SOOOO far apart, out there. In commercial office parks and other such areas, I almost never find open APs that aren't connected to corporate office networks. But the local Starbucks will generally have a T-Mobile, or something.
Of course, if you're willing to use open APs on corporate networks, this changes a lot, because any good-sized office park will have a few rogue APs hooked up. It's an ethical decision for me, based on my experiences with those kind of setups, and I can see others legitimetely feeling differently.
The ideal place is a downtown or hipster neigborhood in a Western city, like San Francisco or Seattle. There are tons of residential APs that are left open, many intentionally, and the signals propagate a good distance away. If you're lucky, you'll probably find a funky coffee shop/bakery/whathaveyou that has a free AP set up for customers.
Also, it's more common for commercial networks to fence you in by firewall all but the obvious ports, but there are easy ways around that with T-Mobile and the like.
Really, my equation is simple: it's too important to me that I have some kind of reliable wireless access, if at all possible, no matter where I happen to sit down and open my laptop. So the $20/mo to T-Mobile, or even $20+$20 for a pair of services, is still cheap to me because it's going to save my ass when I do need it.
At some point, when the prices get low enough and the Linux drivers get good enough, I'm just gonna buy a fucking EVDO card and be done with it, but that's still a little too pricey for me, right now.
I think T-Mobile's rates are just fine, thank you. And it's probably not just because I have a job.
Seriously, one person (or even many people) with the opinion that the pricing is too high for too little doesn't mean that a business model sucks! Last I heard, T-Mobile's hotspot subscriptions were doing pretty damn well. One of the reasons why the hourly rate is so high is to encourage people who use it more than rarely to subscribe, which helps even out the revenue stream and usage patterns. This is just like cellphone billing--plan minutes are loads cheaper per minute than overtime minutes because they want to impose a cost on you for being unpredictable.
As long as they have enough people who pay the freight, bitching or not, nobody else can say that the price is too high to be a "good business model".
Then again, I come from NYC. When I went to LA for the NBA all-star game last year, I remember driving down Figueroa St., about 1/2 block from the Staples Center, and seeing signs for $20 parking spots 1 hour before gametime. I literally said to myself "$20? What a deal! How can these not be taken this late before the game!" Turns out it was because LA people consider $20 for parking to be a ripoff. In NYC, that's kind of a steal.
You're not looking at it from a strategic perspective. Large companies (and smaller ones, too, if they're smart) know that every day they're in business, there's a risk that they're going to get sued, and try to act intelligently about that knowledge. It's called risk management. They also know that they might be able to sue another company/person, too.
This means trying to minimize lawsuit losses (legal costs, judgements against, licensing costs, and bad PR) and maximize gains (judgements for, licensing profits). There's a hell of a lot of thought that goes into this, at most decent-sized corporations, especially when it comes to something like IP lawsuits.
The smart strategy for Microsoft is to weigh the impact that software patents will have on their bottom line, overall. One lost suit (or a dozen) aren't a bad thing, as long as MS makes more money in the process. If they're winning enough suits of enough size, and licensing out their own IP for enough money, the total gains attributable to software patents might outweigh the occasional lawsuit that they lose.
And being that MS has such a mountain of US Patents at its disposal, growing all the time, I think they may be seeing a little further than one or two patent lawsuits, even if they're individually pretty large.
As an aside, most business don't have much lobbying power, and so the point about lobbying for change is nonsense except for the biggest companies, like MS. Even if those few giants are against patents, MS and its ilk will be weighing the effort and expense of lobbying versus the chances that they'll get laws passed--not a sure thing, by any stretch. So even if they suffer some small losses in the long run, it might not be worth the effort to try the change it.
You must be a real dick on long car trips. "Nonono, we can't listen to the radio! All the stations are repetitive ClearChannel crap, and will rot our brains!" If anyone brought any CDs, it's all "That had better not be that mindless dance tune by the teen stat-du-jour."
You know, I listen to KEXP and KCRW by web feed a LOT, because I like the genres they play. And you know what? They're REPETITIVE! That's right, they play the same songs multiple times in a week, sometimes a day! Also, all Indie Rock sounds pretty similar when you listen to it nonstop for a couple of hours. No, they're not more creative because they use a goddamn banjo in a rock song.
Try appreciating some of that pop music for the production values. Brittany Spears is not a particularly distinguished artist, fine. But most of the producers who've worked on her singles are wildly talented musicians with loads of great songs and arrangements to their various credits. Look up Matrix, if you want a good place to start. You'll probably come to realize in this sojourn that the sound of most Indie Rock is ALSO defined heavily by the producer, not the band. That's recorded music for ya!
And while you're at it, take your little-dicked, white-boy head out of your ass and consider how all of that repetitive, uncreative Indie Rock sounds to others... like urban black kids who enjoy rap, or non-urban white kids who like country. Guess what? Their complaints about your favorite crap will sound eeriely similar.
- The root DNS servers know the IP addresses of the DNS servers for.com,.org,.edu, etc
- The.com DNS servers (there are many) know the IP addresses of the DNS servers for google.com, yahoo.com, citibank.com, etc.
- The google.com DNS server knows the IP addresses for the hostnames "www.google.com", "news.google.com", etc.
IF there are any third-level domains under google.com (like corporate.google.com, yomama.google.com, etc) then there are two possibilities:
- the google.com DNS servers know the IP addresses of DNS servers for the subdomains (each subdomain has its own DNS servers), OR
- the google.com DNS server itself also acts as a DNS server for the subdomains, and knows the IP addresses of the hostnames in the subdomains.
Unless you have shitloads of third-level subdomains, and tons of hostnames in those subdomains, it's not generally necessary to run DNS servers for the third-level domains. You could do it, but why make the effort of setting up and maintaining seperate servers for them unless it's necessary to handle a load of traffic?
In theory, you can continue the hierarchy down through as many level of subdomains as you want. I've never personally seen/heard of an independent server for anthing beyond the third level, but they could certainly exist somewhere.
Once you get past the first-level DNS servers (.com,.org, etc), the responsibility of running the DNS servers is in the hands of the domain owner. So Google (the google.com owner) provides IP addresses of its DNS servers to the.com servers, and sets up DNS servers on those IP addresses on its own. In the case of smaller domains, most hosting companies will provide DNS as a part of the hosting package.
Really, the "nutjob" comment wasn't on account of his worries about the dollar's value. Those are fair enough, I'll admit, though the Euro zone has its own set of risks, not the least of which being that the European central banks have been having problems maintaining fiscal restraint.
No, the "nutjob" appelation is strictly for comments like this:
America hasn't has a real currency since leaving the gold standard (1933) and silver standard (1970-71 right?). It's a shame to realize that the largest counterfeit agency is still the federal reserve (which by the way is a private company and NOT a true part of the federal government... much like federal express is not a government agency).
Huh? Calling the Fed the "largest counterfeit agency"? A private company, right, the head of which is appointed by the president of the USA?
The Fed is more like the Post Office or something than the IRS, but that doesn't mean it's not part of the government. Or how about this:
I'm serious people... US paper money isn't real money! I wish I could arrest people for paying me in that worthless crap
Sounds like he's about three standard deviations to the right of Pat Buchanan, and three to the crazy of Sideshow Bob.
Plus, did you check the link in his.sig? That alone is worth a couple of laughs.
Posssibly... POSSIBLY... he's also a clever little troll. The website is just wacky enough.
Ahh, it's so good to read people who understand WTF legal tender actually means. And that's a damn good point about the existing debt versus refusing to accept payment in some form before the transaction--I didn't even think to make the distinction.
"technically correct"? Where in the hell do you get THAT from?!? The guy was accused of using COUNTERFEIT money. What's the definition of counterfeit, now?
The US government issues cash currency. They have laws that say "you're not allowed to try to pass off anything but actual, gov't printed bills as being such." There are lots of specifics to the counterfeiting laws that tell us exactly what it means to try to pass something off like that, if you care to find out. Counterfeit ONLY means that you're passing something off as US-printed money that didn't originate in the usual fashion from the Bureau of Engraving and Printing.
Counterfeiting has nothing to do with the concept of "legal tender", either. Legal tender has to do with what kind of payment of debts you can force on other people. If you owe someone a debt, they *might* accept payment in the form of gold, silver, your used '92 Ford Escort, or cattle... but they HAVE to accept payment in US currency.
Even if the government were to print so much money (or devalue the currency by other means) so as to make dollars essentially worthless for private debt, that doesn't change counterfeiting laws. I'm assuming that currency devaluation is what your particular species of nutjob is worried about, BTW.
***OFFTOPIC RANTING***
Is this all a good idea? Most modern governments seem to think so. Being as gold- or silver-backed currency only has value by mutual agreement, anyway, it probably doesn't matter. The only difference is that it's harder to inflate a currency when it's backed by precious metals... but that doesn't mean it's not possible to do so.
Look at what happened to Europe in the 1500-1600s when massive amounts of new gold bullion came back across the Atlantic from South America. Massive inflation. What would happen if we started mining asteroids, finding enormous amounts of gold, and bringing it back to Earth? Massive inflation. What would happen if huge new gold mines were discovered in an easy-to-access fashion? Same thing. No currency can be fully inflation-proof unless there is actually, without exception, no way to increase the amount of it.
Not that governments don't occasionally bring down a currency and ruin a lot of people by letting inflation run wild, but fiat currency has enough other advantages (the modern banking system, for one) that most informed commentators consider it an acceptable danger.
Sorry for the lesson in modern currency concepts--back to your regularly scheduled Slashdot.
Linux frowns on binary-only, closed-source drivers for a reason: they decrease the overall kernel quality if nobody else can help debug them. Nvidia's closed-source driver is fine, until its doesn't fucking work. Then what?
And this is kernel space we're talking about, so this means that your machine keeps crashing, hard, when it fucks up. And nobody can fix it, except the vendor, who "updates them as frequently as they update the Windows drivers", which means about twice a year, no more than four times total over the life of the product.
This is NOT good enough. The Linux kernel changes much more frequently and drastically than Windows, and driver maintainers are expected to keep up with the kernel or have their code cut out.
Torvalds and the kernel maintainers are driving a very particular type of bus, here. People who want to release binary-only drivers are just unwilling to get on the bus.
It was a mix of the prof and his three TAs, for a class of about 40 kids. They did compile and do an output check, but you could still fail the assignment with zero points if it compiled and worked. Some examples of why:
- no comments in source; - bad variable names confuse reader; - sexual insults leveled at professor in variable names; - wrote source in C++ (the class was on C); - extraordinarily obtuse algorithms that miss the point of the exercise; - calling library functions to perform work that you specifically asked to implement yourself;
Now, to be fair, all of this was laid out in the syllabus and in every single assignment description.
But the point is, they read your source pretty thoroughly and graded the source, not the result. They were trying to teach us how to write good source code.
"Neural nets aren't based on the way animal brains work..."
Jesus H. Christ. GO BACK AND READ MY POST. I didn't say neural nets "are based on animal brains" or "based on the way animal brains work". I made a VERY limited statement about the two: "based on the fundamental theory of how animal brains... operate". That's pretty goddamn different.
And had you bothered to read the OTHER response to this post, you'd see that we got into that already. And established that yes, neural nets and animal brains are different.
The point is that neurons in an animal brain and neural nets build algorithms in the same way, because they process information in the same basic fashion. Animal brains are more complex, true. Animal brains are not completely understood, true. But the same underlying concept of information flow is at work in both.
"That's pretty much unsupported. Spammers do try to trick Bayesian filters, but there's no evidence that they have any success."
Um, I'VE been able to trick Bayesian filters, I spent a couple of hours my coworker's well-trained Thunderbird filter, tweaking spam messages that I received to get by it. Managed to achieve a pretty high success rate, with a few simple rules. I got the idea after reading a Register article about spammers doing the same thing, and I wanted to see how hard it was. Turns out, not hard at all.
So the rest of your commentary that that is just, well, bullshit.
The problem with tricking Bayesian filters is that if someone re-trains the filter to deal with your new tricks, they don't work anymore. Is THAT what you meant? Because you do an awful job of communicating it. Once my coworker tagged a couple dozen of the trick messages as spam, they stopped getting through so reliably. But then I just switched my tricks again, and it worked again. As long as I could keep thinking up new tricks and pushing the envelope, I could get stuff by his filter.
Slashdot Mirror
I don't get it--since when do we tax assets? We don't do that here in the US, and I didn't think that Europe did, either. Keeping 2 years worth of widgets on hand isn't expensive because of taxes, it's expensive because of the storage space it takes up.
But on top of that, companies already DO pay for the government's "protection": they pay corporate income taxes, sales taxes, and more.
And it's kind of silly, from a tax standpoint, to let the individual companies determine the value of their own intellectual property. Once an asset is "on the books", a company can value it however they want, which may result in some absurd claimed valuations depending on how they do their accounting. For instance, the company that owns the Empire State Building in NYC lists the asset on its general ledger as being worth exactly $1. No, NOT $1 million, but one single dollar. They're perfectly allowed to do that, and there is a sensible reason for it as far as how they initially dealt with the costs of building the place.
So if you say "pay %1 maintenance for intellectual property", the big media companies will just adjust their booked asset values downward. You could always pass a law preventing them from doing that, but such a law would have lots of horrible unintended consequences, because the ability to book arbitrary asset values is a necessary part of accounting.
Um, no. MS could easily sell you Windows without asking you to agree with a EULA/license. But they do.
They have the right to forbid others from using their software, because they own the copyright. That's what allows them to demand payment: "If you don't pay us $300, you can't use Windows." They also demand that you agree to the EULA, in order to be allowed to install the software on your machine.
Now, the legality of some of these implementations (click-thru licenses, pre-installed licenses) has been questioned, but the basics are known to be true:
1) If I own the copyright on a work, I can forbid others from using the it.
2) In exchange for being allowed to use my works, I can demand arbitrary concessions from other parties.
No, because unless you agree to a license/contract when you buy it, they can't enforce any terms on you, regardless of the fact that they own a copyright on the software that runs on the camera. You're not allowed to copy the copyrighted software on the camera without their permission, but that's the limit of what they can claim without a license agreement.
My coffee maker has software--it's a real flash job. I'm not allowed to copy that software, but they can't claim that any code I write while drinking the coffee made with it is a "derived work"... Maybe that's a cheeky example.
MS owns the copyright on Windows. They agree to let you use their software IF you agree to their EULA/license. If they just sell you the software without making you agree to the license, they have no power over you other than to prevent you from making unauthorized copies.
Ooh, didn't know that. Well, 10% is something... Maybe I'm wrong about that.
A good comment, but you didn't quite get what I'm asking in the OP.
What the DMCA does prevent is anybody (like Adobe) from "traficking" in software which decrypts the data. It doesn't matter who actually owns the copyright...even if the photographer releases everything to the Public Domain. It's the fact that there could exist any copyrighted photo which Adobe's software could be used to decrypt by anybody else who's not the copyright holder.
THAT'S what the question is: does the DMCA make encryption tools unlawful that are used to decrypt data to which the decrypting party own the copyright? That's not clear at all, to me. Adobe THINKS that it may be the case, and so they've decided to scoot around the issue... But nobody's actually established to me, yet, that this is really the case!
Look at it this way: it's NOT illegal to decrypt any encrypted information. You can write a WEP-cracking tool, or a Kerberos password brute-forcer, or Jack the Ripper, and you don't run afoul of the DMCA. The DMCA only comes into effect when you're talking about breaking encryption to violate copyright.
Adobe could well be acting over-carefully, here, just to avoid the issue entirely.
Sorry, Charlie. You missed this boat. Too bad you didn't RTFA.
in the very quote you provided, Nikon claims to own the copyright on that information
The guy who wrote the PhotoshopNews article made the statement that "...Nikon apparantly thinks they own the information..." (emphasis mine). As far as any of the article says, or any of the links reveal, Nikon hasn't said ANYTHING about this case. Adobe's employees said something, but not Nikon.
The auther is assuming from the Adobe people's statements that they think that Nikon feels a certain way. But Nikon never actually said that. And if Nikon isn't claiming a copyright on the data in the file, they could still encrypt it in order to make 3rd-party software less useful. Nothing stops them from being bastards.
And now that I think about it, why IS the white balance information any different from the image? The image is just bits, which make a recognizable image when interpreted a certain way by image software. The copyrighted form of the image IS the bits that describe it. Some of those bits are encoded in the white balance information--they help describe the image. Just because most cameras don't report the white balance data doesn't mean that it isn't "part of the image."
They likely retain their copyright over the "NEF" part even if it is incorporated into the subsequent photograph.
That doesn't make any sense, though. You don't have a EULA when you buy a camera, because they're not licensing the camera for you to use. They're SELLING you the camera, which you then own. There's no copyright over the photographs that Nikon ever holds that would allow them to "retain" some portion of it.
What you're describing is like a typewriter manufacturer claiming copyright on anything typed on their equipment, or a refrigerator maker claiming copyright on the pictures that my kids draw that I hang on the front of it. Since there's no agreement between me and the manufacturer that I'm assigning copyrights to them as a condition of using their product, they can't possibly own copyrights on the work.
I'm not saying that Nikon COULDN'T have such an agreement with the camera (maybe, maybe not), just that I'm pretty sure no such agreement exists when you buy one.
So it's a way for them to lower the retail price of the camera, similar to the way rebates work.
That doesn't make ANY sense. This is a $5,000+ camera--an extra $100 on the retail price isn't going to have shit to do with anyone's decision to buy it. When you're buying high-end stuff like this, you're probably not making a decision based solely on comparative prices, and you're also probably not THAT sensitive to price that you're care about a 2% difference. When people buy Bentley and Rolls-Royce cars, they don't care that the CD player costs $500 more than an after-market shop would charge, because they're dropping $100K on the car, anyway.
More likely, Nikon wanted to add the raw shooting mode to the camera, requiring white balance data in the file format, but wasn't entirely sure that software on the market (like Photoshop) would properly support the proprietary stuff when the camera hit the market. So they spent the cash to write their own software in-house. Then, somebody says "why are we giving the software away for free?" and they decide to charge for it, which justifies a decision to encrypt the data in the file to force the business model on the consumer.
Which looks like a dumb decision from here, but that's probably because different people made different parts of the decision, probably without stopping to consider the bigger picture: no professional photog is going to want to use their shitty proprietary software, if they can help it.
Not exactly. Assuming that Nikon actually owns the rights to the algorithm that they're using for the encryption (they might not, instead using someone else's or a public domain algorithm), licensing it doesn't get Adobe anywhere. The DMCA doesn't say you're not allowed to use an algorithm, it says you're not allowed to actually decrypt the data.
I think you meant that Adobe should get an agreement with Nikon that they're allowed to build and sell the decryption software, and that Adobe's customers are allowed to use it on their photos. That's not a licensing agreement, exactly... I don't know what you'd call it.
(I personally think that would be a bogus interpretation of the DMCA, since I think the copyrighted information inside the NEF file belongs to the photographer, not Nikon. But Nikon apparently thinks they own the information inside the NEF).
This is a little strange, isn't it? If a photographer takes a picture, it's pretty clear that the photog owns the copyright to that photo. Nikon couldn't possibly claim any rights on photos taken with their camera, least of all because it would make it impossible for professionals to use that equipment. And with $5K cameras, you're really only looking at the professional market.
So if the white balance information (the encrypted stuff) is a part of the photograph, the photographer owns the copyright on that data, too, right? That seems pretty straightforward, but I could be wrong...
Can the DMCA be applied to prevent you from decrypting something that you own the copyright on? This isn't even like owning a DVD and wanting to decrypt the data, because in that case the movie company owns the copyright.
If the DMCA can be applied that way, that's some fucked-up shit. It's just absurd.
RMS may be a little freaky... Okay, he's a LOT freaky. I think that most people are put off more, though, by the near-religious aspects of his approach to free software.
I'm a big free software fan, but I frame it in highly practical terms. Copyright is a monopoly, by definition--to some extent, it's a justifiable monopoly, because of the benefits of guaranteeing an economic return to the parties that create and fund its development. It helps to economically rationalize the costs of software development, because the work of creating it can be rewarded in a way that would otherwise be difficult.
But the downside of the monopoly is pretty big, too--proprietary lock in is the big one, in my mind, but there's more. If you're reliant on closed-source software, you usually have to take it as-it, bugs and missing features and bloat, and all. OSS tends to encourage a lot more flexibility, interoperability, and choice in the way things work. The business benefits of these freedoms are huge, no question.
The great thing about OSS is that it still allows revenue streams for software creators and funders. The creators of MySQL get a lot of consulting business based on the fact that they know the software better than anyone else. People who make big Linux kernel contributions can find jobs easily with companies that need Linux expertise, because they have a proven qualification in that area that means more than a million certifications. So the people that contribute have a payoff.
And GPL-style licenses are more effective at preserving these revenue opportunities than BSD-type licenses, because they tend to prevent proprietary code-forking, which keeps projects under the control of the original creators more easily.
This is the great realization of OSS that can be couched in purely economic terms: all the benefits of copyrightable works, but minimizing the negatives of the monopoly that comes with it.
I don't know how much RMS agrees with this particular sentiment, but then again, I don't really care.
I assume you're talking about the different effects of resistance on AC and DC currents: as electricity travels through a conventional conductor, the resistance of the conductor gives up some of the electrical power as heat (as Ohm's law describes). That's why we use high-voltage AC to distribute electric power, and even higher-voltage AC to transmit power over long distances--by transmitting at high AC voltages, you don't lose quite as much power as you otherwise would.
So if you could replace vast swaths of conventional copper electric transmission and distribution lines with superconductors, you could theoretically switch to DC power in these applications, which would have some interesting effects on the rest of the electrical distribution system.
Strict DC voltage on the power lines would virtually eliminate the EM radiation. You would still get some EM when you turned things on or off, or if the amount of power the line carries changed at all, but there would be a HELL of a lot less.
Lower voltages could be used, which would be safer (less chance of electrocuting people), and connectors (plugs, receptacles) designed with lowe voltages in mind would be cheaper to produce and certify.
Also, many devices in the home (especially computer equipment, or anything with circuit logic in it) need to convert the 110V AC current into much lower-voltage DC (2-5V DC, usually) to operate chip logic. This in generally an inefficient process, with a lot of power given up in the transformers and inverters to heat. Granted, you'd have to redesign all the home devices that currently use AC power directly (mostly lights and appliances) to run on DC, it could be done.
Really, the only problem would be the massive costs of switching over from one standard to another. All of those applicances and such would become useless on the new standard, which means everybody has to go out and buy new stuff. If you tried to switch the distribution over to DC in one go, I can see a lot of people having a lot of problems with it. And it wouldn't be practical to change the distribution bit-by-bit, either.
If you just wanted to change the transmission side, and leave the consumer out of it entirely, you'd have to replace a lot of power generation infrastructure. This could be done more slowly, I'd imagine, but it would still be expensive.
But then again, there's nothing that prevents you from continuing to run AC current on superconducting wires. That's probably what will happen, because it's the cheapest option.
I don't see anyone caring too much about interference from power lines in the 60Hz frequency band, anyway--not like we use those frequencies for anything.
Have you ever heard of a little thing called the "International Space Station"? Yeah, exactly.
If you leave Earth's protective envelope, you do become subject to larger exposures of radiation, but the danger can be minimized with shielding. People can take a decent amount of radiation exposure before they show any health effects, even the long-term cancer risks.
Basically, you just have to build sheltered structures for the inhabitants to block the radiation. It can be a choice of construction materials, or you can just bury the structures and pile regolith on top--the old bomb shelter solution. All it takes is a bulldozer and some internal supports, and you've got a pretty nice shelter.
Spacesuits aren't very good radiation protection, so they'll probably impose daily, monthly, and mission-specific limits on outdoor activities for inhabitants. If they also keep alert to solar activity, they can just head indoors when the worst stuff is coming.
How can you keep alert to incoming radiation, you ask? The ionizing, harmful parts of solar radiation are mostly charged particles, which travel slower than the speed of light. Big emissions of charged particles happen in conjunction with particular types of electromagnetic radition, which DOES travel at the speed of light (duh). So we look for the EM radiation that signals a coming charged particle storm, and tell the moonies to get indoors quick.
Not perfectly safe, but come on: they're on the fucking moon.
The statistics I barely remember from a year ago don't hold much weight, I'll grant you -- but the point is that we don't have to trade anecdotes, we can look at actual case studies. My money is on free wifi translating to better business, and for-pay wifi adding expenses that significantly limit its profitability. In that context, the GP is correct -- wifi providers would do well to get their minds right.
Seems like a lot of effort, given that I've gotSo let me start off by saying "There is a very large chance that you may be correct".
The problem with the OP's post wasn't about whether WiFi providers/locales have chosen the best business model: it's whether one person's assessment that a service costs too much for too little is a credible proof of whether the pricing model is good. That's an absurd conclusion. the fact that a Rolls-Royce limo is too expensive for me to consider buying one has NOTHING to do with the price of tea in China.
I use a LOT of free hotspots--tiny, 2-person startup company, no real office at the moment, so we work out of coffee shops and bars a lot.
Commercial hotspots have one thing that free nodes can't beat: quality of service. In terms of distance to the AP and interference from intervening objects, you'll usually get a much better signal from the commercial AP. This is exacerbated in a built-up city like NYC, because the buildings severely cut down on wireless propagation between blocks--normally, you'll only get strong enough signals to connect to something on the same block, with a rough line-of-sight to the AP.
But I can see that being a non-issue in suburban areas, or the cities out West where the buildings are smaller, fewer, shorter, and lighter of construction. I've noticed in other areas that the propagation issues are balanced by the fact that shit is SOOOO far apart, out there. In commercial office parks and other such areas, I almost never find open APs that aren't connected to corporate office networks. But the local Starbucks will generally have a T-Mobile, or something.
Of course, if you're willing to use open APs on corporate networks, this changes a lot, because any good-sized office park will have a few rogue APs hooked up. It's an ethical decision for me, based on my experiences with those kind of setups, and I can see others legitimetely feeling differently.
The ideal place is a downtown or hipster neigborhood in a Western city, like San Francisco or Seattle. There are tons of residential APs that are left open, many intentionally, and the signals propagate a good distance away. If you're lucky, you'll probably find a funky coffee shop/bakery/whathaveyou that has a free AP set up for customers.
Also, it's more common for commercial networks to fence you in by firewall all but the obvious ports, but there are easy ways around that with T-Mobile and the like.
Really, my equation is simple: it's too important to me that I have some kind of reliable wireless access, if at all possible, no matter where I happen to sit down and open my laptop. So the $20/mo to T-Mobile, or even $20+$20 for a pair of services, is still cheap to me because it's going to save my ass when I do need it.
At some point, when the prices get low enough and the Linux drivers get good enough, I'm just gonna buy a fucking EVDO card and be done with it, but that's still a little too pricey for me, right now.
For another perspective...
I think T-Mobile's rates are just fine, thank you. And it's probably not just because I have a job.
Seriously, one person (or even many people) with the opinion that the pricing is too high for too little doesn't mean that a business model sucks! Last I heard, T-Mobile's hotspot subscriptions were doing pretty damn well. One of the reasons why the hourly rate is so high is to encourage people who use it more than rarely to subscribe, which helps even out the revenue stream and usage patterns. This is just like cellphone billing--plan minutes are loads cheaper per minute than overtime minutes because they want to impose a cost on you for being unpredictable.
As long as they have enough people who pay the freight, bitching or not, nobody else can say that the price is too high to be a "good business model".
Then again, I come from NYC. When I went to LA for the NBA all-star game last year, I remember driving down Figueroa St., about 1/2 block from the Staples Center, and seeing signs for $20 parking spots 1 hour before gametime. I literally said to myself "$20? What a deal! How can these not be taken this late before the game!" Turns out it was because LA people consider $20 for parking to be a ripoff. In NYC, that's kind of a steal.
You're not looking at it from a strategic perspective. Large companies (and smaller ones, too, if they're smart) know that every day they're in business, there's a risk that they're going to get sued, and try to act intelligently about that knowledge. It's called risk management. They also know that they might be able to sue another company/person, too.
This means trying to minimize lawsuit losses (legal costs, judgements against, licensing costs, and bad PR) and maximize gains (judgements for, licensing profits). There's a hell of a lot of thought that goes into this, at most decent-sized corporations, especially when it comes to something like IP lawsuits.
The smart strategy for Microsoft is to weigh the impact that software patents will have on their bottom line, overall. One lost suit (or a dozen) aren't a bad thing, as long as MS makes more money in the process. If they're winning enough suits of enough size, and licensing out their own IP for enough money, the total gains attributable to software patents might outweigh the occasional lawsuit that they lose.
And being that MS has such a mountain of US Patents at its disposal, growing all the time, I think they may be seeing a little further than one or two patent lawsuits, even if they're individually pretty large.
As an aside, most business don't have much lobbying power, and so the point about lobbying for change is nonsense except for the biggest companies, like MS. Even if those few giants are against patents, MS and its ilk will be weighing the effort and expense of lobbying versus the chances that they'll get laws passed--not a sure thing, by any stretch. So even if they suffer some small losses in the long run, it might not be worth the effort to try the change it.
You must be a real dick on long car trips. "Nonono, we can't listen to the radio! All the stations are repetitive ClearChannel crap, and will rot our brains!" If anyone brought any CDs, it's all "That had better not be that mindless dance tune by the teen stat-du-jour."
You know, I listen to KEXP and KCRW by web feed a LOT, because I like the genres they play. And you know what? They're REPETITIVE! That's right, they play the same songs multiple times in a week, sometimes a day! Also, all Indie Rock sounds pretty similar when you listen to it nonstop for a couple of hours. No, they're not more creative because they use a goddamn banjo in a rock song.
Try appreciating some of that pop music for the production values. Brittany Spears is not a particularly distinguished artist, fine. But most of the producers who've worked on her singles are wildly talented musicians with loads of great songs and arrangements to their various credits. Look up Matrix, if you want a good place to start. You'll probably come to realize in this sojourn that the sound of most Indie Rock is ALSO defined heavily by the producer, not the band. That's recorded music for ya!
And while you're at it, take your little-dicked, white-boy head out of your ass and consider how all of that repetitive, uncreative Indie Rock sounds to others... like urban black kids who enjoy rap, or non-urban white kids who like country. Guess what? Their complaints about your favorite crap will sound eeriely similar.
DNS is a hierarchy of names:
.com, .org, .edu, etc .com DNS servers (there are many) know the IP addresses of the DNS servers for google.com, yahoo.com, citibank.com, etc.
.org, etc), the responsibility of running the DNS servers is in the hands of the domain owner. So Google (the google.com owner) provides IP addresses of its DNS servers to the .com servers, and sets up DNS servers on those IP addresses on its own. In the case of smaller domains, most hosting companies will provide DNS as a part of the hosting package.
- The root DNS servers know the IP addresses of the DNS servers for
- The
- The google.com DNS server knows the IP addresses for the hostnames "www.google.com", "news.google.com", etc.
IF there are any third-level domains under google.com (like corporate.google.com, yomama.google.com, etc) then there are two possibilities:
- the google.com DNS servers know the IP addresses of DNS servers for the subdomains (each subdomain has its own DNS servers), OR
- the google.com DNS server itself also acts as a DNS server for the subdomains, and knows the IP addresses of the hostnames in the subdomains.
Unless you have shitloads of third-level subdomains, and tons of hostnames in those subdomains, it's not generally necessary to run DNS servers for the third-level domains. You could do it, but why make the effort of setting up and maintaining seperate servers for them unless it's necessary to handle a load of traffic?
In theory, you can continue the hierarchy down through as many level of subdomains as you want. I've never personally seen/heard of an independent server for anthing beyond the third level, but they could certainly exist somewhere.
Once you get past the first-level DNS servers (.com,
Really, the "nutjob" comment wasn't on account of his worries about the dollar's value. Those are fair enough, I'll admit, though the Euro zone has its own set of risks, not the least of which being that the European central banks have been having problems maintaining fiscal restraint.
.sig? That alone is worth a couple of laughs.
No, the "nutjob" appelation is strictly for comments like this:
America hasn't has a real currency since leaving the gold standard (1933) and silver standard (1970-71 right?). It's a shame to realize that the largest counterfeit agency is still the federal reserve (which by the way is a private company and NOT a true part of the federal government... much like federal express is not a government agency).
Huh? Calling the Fed the "largest counterfeit agency"? A private company, right, the head of which is appointed by the president of the USA?
The Fed is more like the Post Office or something than the IRS, but that doesn't mean it's not part of the government. Or how about this:
I'm serious people... US paper money isn't real money! I wish I could arrest people for paying me in that worthless crap
Sounds like he's about three standard deviations to the right of Pat Buchanan, and three to the crazy of Sideshow Bob.
Plus, did you check the link in his
Posssibly... POSSIBLY... he's also a clever little troll. The website is just wacky enough.
Ahh, it's so good to read people who understand WTF legal tender actually means. And that's a damn good point about the existing debt versus refusing to accept payment in some form before the transaction--I didn't even think to make the distinction.
"technically correct"? Where in the hell do you get THAT from?!? The guy was accused of using COUNTERFEIT money. What's the definition of counterfeit, now?
The US government issues cash currency. They have laws that say "you're not allowed to try to pass off anything but actual, gov't printed bills as being such." There are lots of specifics to the counterfeiting laws that tell us exactly what it means to try to pass something off like that, if you care to find out. Counterfeit ONLY means that you're passing something off as US-printed money that didn't originate in the usual fashion from the Bureau of Engraving and Printing.
Counterfeiting has nothing to do with the concept of "legal tender", either. Legal tender has to do with what kind of payment of debts you can force on other people. If you owe someone a debt, they *might* accept payment in the form of gold, silver, your used '92 Ford Escort, or cattle... but they HAVE to accept payment in US currency.
Even if the government were to print so much money (or devalue the currency by other means) so as to make dollars essentially worthless for private debt, that doesn't change counterfeiting laws. I'm assuming that currency devaluation is what your particular species of nutjob is worried about, BTW.
***OFFTOPIC RANTING***
Is this all a good idea? Most modern governments seem to think so. Being as gold- or silver-backed currency only has value by mutual agreement, anyway, it probably doesn't matter. The only difference is that it's harder to inflate a currency when it's backed by precious metals... but that doesn't mean it's not possible to do so.
Look at what happened to Europe in the 1500-1600s when massive amounts of new gold bullion came back across the Atlantic from South America. Massive inflation. What would happen if we started mining asteroids, finding enormous amounts of gold, and bringing it back to Earth? Massive inflation. What would happen if huge new gold mines were discovered in an easy-to-access fashion? Same thing. No currency can be fully inflation-proof unless there is actually, without exception, no way to increase the amount of it.
Not that governments don't occasionally bring down a currency and ruin a lot of people by letting inflation run wild, but fiat currency has enough other advantages (the modern banking system, for one) that most informed commentators consider it an acceptable danger.
Sorry for the lesson in modern currency concepts--back to your regularly scheduled Slashdot.
Linux frowns on binary-only, closed-source drivers for a reason: they decrease the overall kernel quality if nobody else can help debug them. Nvidia's closed-source driver is fine, until its doesn't fucking work. Then what?
And this is kernel space we're talking about, so this means that your machine keeps crashing, hard, when it fucks up. And nobody can fix it, except the vendor, who "updates them as frequently as they update the Windows drivers", which means about twice a year, no more than four times total over the life of the product.
This is NOT good enough. The Linux kernel changes much more frequently and drastically than Windows, and driver maintainers are expected to keep up with the kernel or have their code cut out.
Torvalds and the kernel maintainers are driving a very particular type of bus, here. People who want to release binary-only drivers are just unwilling to get on the bus.
It was a mix of the prof and his three TAs, for a class of about 40 kids. They did compile and do an output check, but you could still fail the assignment with zero points if it compiled and worked. Some examples of why:
- no comments in source;
- bad variable names confuse reader;
- sexual insults leveled at professor in variable names;
- wrote source in C++ (the class was on C);
- extraordinarily obtuse algorithms that miss the point of the exercise;
- calling library functions to perform work that you specifically asked to implement yourself;
Now, to be fair, all of this was laid out in the syllabus and in every single assignment description.
But the point is, they read your source pretty thoroughly and graded the source, not the result. They were trying to teach us how to write good source code.
"Neural nets aren't based on the way animal brains work..."
Jesus H. Christ. GO BACK AND READ MY POST. I didn't say neural nets "are based on animal brains" or "based on the way animal brains work". I made a VERY limited statement about the two: "based on the fundamental theory of how animal brains... operate". That's pretty goddamn different.
And had you bothered to read the OTHER response to this post, you'd see that we got into that already. And established that yes, neural nets and animal brains are different.
The point is that neurons in an animal brain and neural nets build algorithms in the same way, because they process information in the same basic fashion. Animal brains are more complex, true. Animal brains are not completely understood, true. But the same underlying concept of information flow is at work in both.
"That's pretty much unsupported. Spammers do try to trick Bayesian filters, but there's no evidence that they have any success."
Um, I'VE been able to trick Bayesian filters, I spent a couple of hours my coworker's well-trained Thunderbird filter, tweaking spam messages that I received to get by it. Managed to achieve a pretty high success rate, with a few simple rules. I got the idea after reading a Register article about spammers doing the same thing, and I wanted to see how hard it was. Turns out, not hard at all.
So the rest of your commentary that that is just, well, bullshit.
The problem with tricking Bayesian filters is that if someone re-trains the filter to deal with your new tricks, they don't work anymore. Is THAT what you meant? Because you do an awful job of communicating it. Once my coworker tagged a couple dozen of the trick messages as spam, they stopped getting through so reliably. But then I just switched my tricks again, and it worked again. As long as I could keep thinking up new tricks and pushing the envelope, I could get stuff by his filter.