Yeah, but it definitely increases the risk that one compromised box will allow access to multiple others.
Consider for example that you may not immediately know when your box is compromised, your users may ignore your suggestions, or that the system compromised may be run by a sys admin who doesn't warn his users (who may now have compromised passwords on your system, even though your system security is fine).
Although I agree that you can imagine far worse things than obtaining user account passwords happening anytime a root account is cracked.
Interesting to hear from the author, but it highlights one of the problems with the article.
That is, how can he claim a result is "skewed"? AFAIK, there isn't some objectively ordered list of web sites--there's always going to be some criteria it's based on, which will always suit some people better than others. What's so wrong with Google's mean of ranking things?
There seems to be an ideal mix in his head that he never bothers to define, but he is willing to criticize Google for deviating from it. For the argument to have any logical basis, he'd at least need to spell out an "ideal," and present a case that his imagined mix is better than Google's current lists.
Fair enough, it does meet the letter of what I proposed, but not quite on the same scale. These seem more standard filing fees than disincentives; you're getting 20 years of patent protection for less than the cost of registering a car over the same period.
I was imagining fees in the six or low seven figure range.
An old idea to acomplish this 'expiration' (a version is in Lessig's Future of Ideas but it's been around awhile) is that individuals have to refile every five years or so to extend patents up to the 20 year point, with a non-trivial, increasing fee associated.
The theory is this would prevent many patents whose purpose or effect is to penalize other's innovations, such as the ones described in this article. Have a half-formed idea but not going to develop it? Can't overcome the technical roadblocks yourself? Just hoping to ambush to some company that has independently done all the hard work re-inventing and then actually developing your idea? Not quite so attractive if you're coughing up out of pocket while you wait around.
When I got my laptop this year, I installed a (legitimate) copy of MS Word 95 I had, because my Word 97 had an OEM license. I thought this made me overly anal about licenses. Apparently, according to the BSA, this makes me a pirate, as I had a demand for the software this year, but didn't make a purchase.
The logic is really too easy to mock. A Hollywood movie flops? The movie quality isn't low, it's just a lot of people must have snuck in--after all, ticket sales didn't match demand estimates. Levi's sales troubles? It's clearly those Hong Kong knockoffs. Any failed product becomes, by definition, a bust due to counterfeiting or theft, not bad projections, marketing or design.
A company that publishes its own trade secrets is pretty much hosed regardless of the license, isn't it? They're required to make reasonable efforts to keep them confidential, which presumably doesn't include selling them publicly. I don't think the "we didn't know about it" argument would hold up--if they want to keep it secret, it's the company's responsibility to know about it.
Personally I think that the courts are the place to argue out the rules of innovation as if you believe in the idea strongly enough then you will be willing to fight, or raise finance to do so. If this forms part of your business proposal then that is right and good. Business decisions are implicitly risky and this will have to be bourne in mind.
The problem is this puts an immense barrier in front of new ideas. If their legality is not clear, and a lawsuit costs hundreds of thousands of dollars to fight (and potentially far more if you lose), the only ideas you'll see tried are ones that are likely to make the innovator millions. What happens to experimentation with novel ideas, ones which you have no confidence up front will work? Should they just be abandoned?
Clarity in the law, so people know what they can try, is far preferable to a patchwork of court decisions.
"We might create a group of people much smarter than us, that might want to kill us," said bioethicist George Annas, chair of the Health Law Department of Boston University School of Public Health.
Or they might be so much smarter than us that they realize they don't need to kill everyone who differs from them . ..
(Comment borrowed from Sladek's "Roderick at Random")
Drug companies dream of preventive medicines, which they could give to all patients, before they actually got sick. It's a much larger market than sick people. Look at the huge sums spent on researching cholesterol-lowering drugs.
In practice, the problem is the minute risk associated with taking any drug is often acceptable when all it's doing is reducing a minute risk of getting sick in the first place.
So drug companies research the drugs that people would actually use (ie: pay for) instead. Go figure.
No. In general, government (and UCSD is a state school) can't discriminate on the grounds they find something objectional or against "their" beliefs. In this case, if they offer web hosting services to all their students, they couldn't then go through and revoke it for those expressing selected views, and a policy claiming they could would be illegal.
People who say the school is being forced to do this by the Patriot act may have this backwards--they may have been looking for an excuse to take it down, and that was the best one.
I would have considered it a proper debunking if it had done a peoper statistical analysis of the deaths
Well, no one can claim the article is rigorously scientific, but I think the valid point is that you can't do a "proper statistical analysis." Or more precisely, you can, but it doesn't tell you anything. Someone could precisley calculate the odds at 20 to one, or 20 million to one, against the murders happening by chance--so what? Neither way does it tell you if there was some common cause. Satistics merely quantify what we don't know. The only way to tell if there's a realted cause is through investigations, which AFAIK haven't found evidence, only non-causal similarities.
Slashdot Mirror
Yeah, but it definitely increases the risk that one compromised box will allow access to multiple others.
Consider for example that you may not immediately know when your box is compromised, your users may ignore your suggestions, or that the system compromised may be run by a sys admin who doesn't warn his users (who may now have compromised passwords on your system, even though your system security is fine).
Although I agree that you can imagine far worse things than obtaining user account passwords happening anytime a root account is cracked.
Interesting to hear from the author, but it highlights one of the problems with the article.
That is, how can he claim a result is "skewed"? AFAIK, there isn't some objectively ordered list of web sites--there's always going to be some criteria it's based on, which will always suit some people better than others. What's so wrong with Google's mean of ranking things?
There seems to be an ideal mix in his head that he never bothers to define, but he is willing to criticize Google for deviating from it. For the argument to have any logical basis, he'd at least need to spell out an "ideal," and present a case that his imagined mix is better than Google's current lists.
Fair enough, it does meet the letter of what I proposed, but not quite on the same scale. These seem more standard filing fees than disincentives; you're getting 20 years of patent protection for less than the cost of registering a car over the same period.
I was imagining fees in the six or low seven figure range.
An old idea to acomplish this 'expiration' (a version is in Lessig's Future of Ideas but it's been around awhile) is that individuals have to refile every five years or so to extend patents up to the 20 year point, with a non-trivial, increasing fee associated.
The theory is this would prevent many patents whose purpose or effect is to penalize other's innovations, such as the ones described in this article. Have a half-formed idea but not going to develop it? Can't overcome the technical roadblocks yourself? Just hoping to ambush to some company that has independently done all the hard work re-inventing and then actually developing your idea? Not quite so attractive if you're coughing up out of pocket while you wait around.
When I got my laptop this year, I installed a (legitimate) copy of MS Word 95 I had, because my Word 97 had an OEM license. I thought this made me overly anal about licenses. Apparently, according to the BSA, this makes me a pirate, as I had a demand for the software this year, but didn't make a purchase.
The logic is really too easy to mock. A Hollywood movie flops? The movie quality isn't low, it's just a lot of people must have snuck in--after all, ticket sales didn't match demand estimates. Levi's sales troubles? It's clearly those Hong Kong knockoffs. Any failed product becomes, by definition, a bust due to counterfeiting or theft, not bad projections, marketing or design.
Decimate can be legitimately used to mean "kill in large numbers" today, regardless of the word's origin. A check of any dictionary will confirm this.
A word's etymology is not the same as its definition.
A company that publishes its own trade secrets is pretty much hosed regardless of the license, isn't it? They're required to make reasonable efforts to keep them confidential, which presumably doesn't include selling them publicly. I don't think the "we didn't know about it" argument would hold up--if they want to keep it secret, it's the company's responsibility to know about it.
Personally I think that the courts are the place to argue out the rules of innovation as if you believe in the idea strongly enough then you will be willing to fight, or raise finance to do so. If this forms part of your business proposal then that is right and good. Business decisions are implicitly risky and this will have to be bourne in mind.
The problem is this puts an immense barrier in front of new ideas. If their legality is not clear, and a lawsuit costs hundreds of thousands of dollars to fight (and potentially far more if you lose), the only ideas you'll see tried are ones that are likely to make the innovator millions. What happens to experimentation with novel ideas, ones which you have no confidence up front will work? Should they just be abandoned?
Clarity in the law, so people know what they can try, is far preferable to a patchwork of court decisions.
"We might create a group of people much smarter than us, that might want to kill us," said bioethicist George Annas, chair of the Health Law Department of Boston University School of Public Health.
.
Or they might be so much smarter than us that they realize they don't need to kill everyone who differs from them . .
(Comment borrowed from Sladek's "Roderick at Random")
Hogwash.
Drug companies dream of preventive medicines, which they could give to all patients, before they actually got sick. It's a much larger market than sick people. Look at the huge sums spent on researching cholesterol-lowering drugs.
In practice, the problem is the minute risk associated with taking any drug is often acceptable when all it's doing is reducing a minute risk of getting sick in the first place.
So drug companies research the drugs that people would actually use (ie: pay for) instead. Go figure.
No. In general, government (and UCSD is a state school) can't discriminate on the grounds they find something objectional or against "their" beliefs. In this case, if they offer web hosting services to all their students, they couldn't then go through and revoke it for those expressing selected views, and a policy claiming they could would be illegal.
People who say the school is being forced to do this by the Patriot act may have this backwards--they may have been looking for an excuse to take it down, and that was the best one.
I would have considered it a proper debunking if it had done a peoper statistical analysis of the deaths
Well, no one can claim the article is rigorously scientific, but I think the valid point is that you can't do a "proper statistical analysis." Or more precisely, you can, but it doesn't tell you anything. Someone could precisley calculate the odds at 20 to one, or 20 million to one, against the murders happening by chance--so what? Neither way does it tell you if there was some common cause. Satistics merely quantify what we don't know. The only way to tell if there's a realted cause is through investigations, which AFAIK haven't found evidence, only non-causal similarities.
Freeman Dyson, who worked on the original Orion project, now estimates each launch would kill one person. He's not so hot on the idea any more.
Obviously, if the Footfall aliens attack, the cost-benefit calculation changes slightly.