I'm considering publishing a list of the IPs I block, and my reasons for doing so: as others weigh in (agreeing or dissenting), it could become the ultimate democracy...
Yeah, the "ultimate democracy." Where despotic regimes harbor cyber miscreants who piss off the inhabitants of "civilized" countries, who block those despotic regimes, therefore denying the innocent inhabitants of those regimes the ability to communicate unfettered with the rest of the free world.
"Hey, there seem to be all these hackers in China. Let's block the entire nation of China from the rest of the Internet. That will really help the Chinese Internet censorship situation."
But I guess your own convenience is more important that giving those people a conduit to freedom.
As somebody else pointed out, an individual has every right to block or receive whatever traffic they wish. But if you're a network administrator at an ISP or government who thinks he's doing some good by closing off these segments of the Internet, you're nothing but low life scum who cares more about his temporary comfort that other people's lives.
For those of us that care about English this is as jarring a syntax error
It's a LEXICAL error. The person understands the difference between the two, he/she just can't spell it correctly. If you're going to make a comparison with compilers, get the fucking terminology correct.
Anyway... Are you trying to say that you're only as smart as a compiler? Yeah, confusing the two is incorrect, but if your mental parser can't deal with a simple misplaced apostrophe... are you saying that's a sign that you're intelligent?
I think mother earth just begins to show us how she likes to be changed by humans.
Drop the "Mother Earth" bullshit. What's happening is that a system is reacting to forcing. If you push a swing at the right frequency, hard enough, you can cause your two year old to go flying off the seat. That doesn't mean the swing "doesn't like being pushed that hard."
This question is somewhat morbid, but I can't help but ask it...
Could "the terrorists" (or really, anybody evil enough) have damaged the levees to the same extent that the hurricane did? At least with a hurricane, you have days of warning to evacuate. Having the levees explode suddenly would probably drown tens or hundreds of thousands of people. It would be an utter catastrophe.
What sort of security is (now, was) there around the levees? And how feasible would it have been for a relatively small team of people to plant enough explosives to blow open a big enough hole?
I've read that one of the breaches is 200 feet wide. Did it start out smaller and then expand as water rushed through it?
If all the buildings were up to code, there wouldn't have been anything but extremely isolated structural damage. But you don't listen. So you die.
Nice, you're saying that the people struck by disaster should be allowed to die? I didn't realize we were a nation of barbarians. Thanks for pointing that out.
Maybe people, in the future, should pay for their own risks, the same way the rest of us do. Just a crazy right wing idea.
So you're advocating an "every man for himself" United States? I don't think we can call ourselves a civilized society, if people suffer and die when there are resources that could help them out.
Going back to disaster prone areas again and again may be stupid, but I don't think the answer is to allow the victims to wallow and die. What the fuck sort of civilized country allows that?
True. But only a moron builds on a fault. Or in the immediate danger area of a volcano. Or in a flood plain. Or really needs a study to figure out what will happen if it lets loose:)
You don't have to be a moron, just ignorant. The city of Tacoma in Washington is built on the deposits of massive lahars from Mt. Rainier -- lahars that flowed very recently, in fact. Tacoma can and will be buried tens of feet in something similar to natural concrete, one day.
It's not because the founders of Tacoma were morons, it's because they didn't know about this stuff when they built the city.
That will never happen in this day and age, given current security concerns. Two words: "car bomb."
Yeah that makes sense, a car bomb is WAY more destructive than a Category 4 hurricane.
There are already buildings throughout the world with parking on the ground floor. Why would a prevalence of them make them more attractive as a target? Does building more skyscrapers increase the chance of another 9/11? It just doesn't follow.
Now I can talk to my tv at 100mpbs I can change channels much quicker and theoretically watch more channels at once.
But now you've got enough bandwidth to send video from the television back to the remote. No more of that stupid "picture in picture" crap, you can browse other channels on your remote while watching something else on the big screen.
Whether or not you actually are liable, you can (and may well) be sued for libel and defamation of character. Even if you're innocent, the question arises: how willing are you to defend yourself? How many lawyers will you hire for how many years just to prove you didn't do it?
Depending on your state, you can file a motion under SLAPP laws to protect yourself. If you are innocent of libel/slander, it is not allowable for a larger entity to quash your speech by threatening an expensive legal action.
I find the mouse gestures on Opera are just so much better than those available (with an extension) on Firefox or Mozilla. It's like the difference between day and night.
I like Opera's gestures too, but have you tried the RadialContext plugin for Firefox (just Google it, I won't bother to link)? It's a lot like gestures, and having used it for almost a year now, I find it totally natural.
I switch browsers a lot -- I use Opera's gestures when it's around, Firefox with RadialContext when it's not, and CocoaGestures/Safari when I'm at home using my Mac Mini. It's kind of like being able to use both Emacs and vi comfortably -- totally possible to learn, and it makes you a lot more flexible.
(Speaking of Emacs and vi, have you ever tried running vi inside Emacs? Now there's a mindfuck.)
In academic machine learning, where you're trying to train a system to distinguish between chairs and doorknobs, there's at least a consensus on what chairs and doorknobs are. Spam is a moving target that's continually updated to defeat new filtering methods.
So, you continually test. Nobody said this was easy.
The hard part, which you sort of elide over, is selecting the test dataset. Whose e-mail?
You're right, sorry. I left something out.
You can also do a P-set M-way N-fold cross validation, which is basically the procedure I outlined above, but repeated P times with P different data sets. The mechanics of computing the WLT table and the T-tests remain the same but obviously there are P times as many results. Now, you can simply start plugging in datasets until everybody is satisfied that the test is fair.
Although there don't seem to be any standards in the Open Source community for this, there are definitely standards in the academic community. Spam filters are a subset of machine learners, and there are very specific and well accepted ways of comparing machine learners.
Typically what is done is to select a range of filters/learners that you want to evaluate. A test dataset is also selected (in this case, it would be an archive of spam and nonspam messages, correctly classified). An M-way N-fold cross validation is performed. What this means is that the data set is split into N parts, and N runs are conducted for each classifier, training using N-1 of the parts. The remaining part is used to test the learner. This is repeated, each time holding out a different part of the test set.
This ENTIRE procedure is repeated M times. This gives, ultimately, M*N results. Each column pair of results from a specific pair of learners has a T-test applied to it. This tells the statistical significance of variations in performance. Usually, a 5% or 1% threshold of significance is used.
Once that is completed, something called a WLT table is computed. Each time a learner defeats another learner on a given test, its W ("Win") counter is incremented. Likewise, when a learner loses, the L ("Loss") counter is incremented. When two learners tie (i.e., when the variation is not statistically significant), the T ("Tie") counter is incremented.
The overall "winner" of the comparison is the learner with the maximum value of W-L.
This sounds complicated and bloated, but it is, in fact, how machine learners are tested in academia. The cross validation method, along with checks for statistical significance, is critical to achieving a valid comparison. Simply running the tests once and saying "This filter got 98% correct, and this other filter got 95% correct -- therefore the first filter is better" is NOT sufficient.
ever heard of evolution? corn and tomato have common ancestors, just like sheep and wolf, or corn and sheep.
Of course, and I addressed that in my post. Evolution happens via mutation. Selective breeding might take advantage of mutations when they occur, but it doesn't cause them. No ancient human could have ever bred a tomato into a corn plant.
actually agriculture is nothing but selecting food crops based on various genetic qualities. we've been genetically engineering foods for tens of thousands of years.
Selecting the most desirable individuals and breeding them is not the same thing as genetic engineering. In selective breeding, the total gene pool remains fixed. You don't create new genes by breeding different plants together, you just create new combinations of existing genes. Every once in a while a mutation occurs, but we have no control over it. Genetic engineering is an entirely different thing -- we have learned how to modify, remove, and insert completely new genes into organisms, at will.
Given enough time, could you ever breed corn such that it became a tomato? It would never happen. Tomatos have genes that corn simply doesn't have. No amount of selection is going to magically turn corn into a tomato. But with our modern technology, we are able to take genes from the tomato and insert them into corn. So genetic engineering is really a vastly different thing than breeding domesticated crops.
just like dogs. do you love your dog? you're dog is a genetically modified wolf, warped by mankind into something wholly unnatural.
Once again, no. The genes of a domestic dog are indistinguishable from the genes of wolves. They simply happen to occur in an extremely unlikely combination, influenced by human breeders. It will never be possible, through breeding, to turn a dog into a sheep. But genetic engineering might one day make that possible.
What we are doing genetically to our food is absolutely nothing like selective breeding.
There is no "limit" to how much moisture the air can "hold." Water vapor and air are infinitely miscible, just like alcohol and water. The relative humidity can be far higher than 100% in certain situations.
How do you decide which comes first? How about whichever section of code is bigger?
Now that's a stupid way of looking at it. One line of code which executes billions of times is more important than a million lines which never execute.
And almost all of your routine interaction with the operating system is GNU-centric. If you switched the kernel to something else, it would still be fundamentally the same experience.
It's not GNU-centric, it's UNIX centric. And have you heard? GNU's not UNIX!
What everybody else is saying is right. Use a body-kinetic dynamo, like a hand cranked or pumped device. There are several examples to be found on ThinkGeek.
Even if you were using the best photovoltaic material available, you probably couldn't get more than 20 or 30 watts from a module small enough to carry around with you, and that's only when the sun is shining at its peak.
No matter what your solution, bear in mind that you will additionally need a charge controller and a battery to dump the energy into. You can't just plug a random power source directly into a device which expects clean DC at a specific voltage.
Hence I have explained the emotion as being the logical consequence the circumstances.
What logical thought occurs in the brain of an MDMA user that leads to the emotional state? There is none, it is the chemical which causes it. What you're saying is, "Effects have causes," well, duh. This has nothing to do with logic.
Slashdot Mirror
That's a FANTASTIC idea (no sarcasm), and if I believed in patents I would urge you to patent it ;-)
Yeah, the "ultimate democracy." Where despotic regimes harbor cyber miscreants who piss off the inhabitants of "civilized" countries, who block those despotic regimes, therefore denying the innocent inhabitants of those regimes the ability to communicate unfettered with the rest of the free world.
"Hey, there seem to be all these hackers in China. Let's block the entire nation of China from the rest of the Internet. That will really help the Chinese Internet censorship situation."
But I guess your own convenience is more important that giving those people a conduit to freedom.
As somebody else pointed out, an individual has every right to block or receive whatever traffic they wish. But if you're a network administrator at an ISP or government who thinks he's doing some good by closing off these segments of the Internet, you're nothing but low life scum who cares more about his temporary comfort that other people's lives.
You're right that at low pressures, water is a vapor at low temperatures... But the Ideal Gas Law has absolutely NOTHING to do with it.
It's a LEXICAL error. The person understands the difference between the two, he/she just can't spell it correctly. If you're going to make a comparison with compilers, get the fucking terminology correct.
Anyway... Are you trying to say that you're only as smart as a compiler? Yeah, confusing the two is incorrect, but if your mental parser can't deal with a simple misplaced apostrophe... are you saying that's a sign that you're intelligent?
Drop the "Mother Earth" bullshit. What's happening is that a system is reacting to forcing. If you push a swing at the right frequency, hard enough, you can cause your two year old to go flying off the seat. That doesn't mean the swing "doesn't like being pushed that hard."
Because ONE place isn't experiencing warming? I make a decent income. Therefore, poverty is a myth.
The reason that Katrina reached Category 5 was the mass of abnormally warm water in the Gulf.
What sort of security is (now, was) there around the levees? And how feasible would it have been for a relatively small team of people to plant enough explosives to blow open a big enough hole?
I've read that one of the breaches is 200 feet wide. Did it start out smaller and then expand as water rushed through it?
Nice, you're saying that the people struck by disaster should be allowed to die? I didn't realize we were a nation of barbarians. Thanks for pointing that out.
So you're advocating an "every man for himself" United States? I don't think we can call ourselves a civilized society, if people suffer and die when there are resources that could help them out.
Going back to disaster prone areas again and again may be stupid, but I don't think the answer is to allow the victims to wallow and die. What the fuck sort of civilized country allows that?
You don't have to be a moron, just ignorant. The city of Tacoma in Washington is built on the deposits of massive lahars from Mt. Rainier -- lahars that flowed very recently, in fact. Tacoma can and will be buried tens of feet in something similar to natural concrete, one day.
It's not because the founders of Tacoma were morons, it's because they didn't know about this stuff when they built the city.
Yeah that makes sense, a car bomb is WAY more destructive than a Category 4 hurricane.
There are already buildings throughout the world with parking on the ground floor. Why would a prevalence of them make them more attractive as a target? Does building more skyscrapers increase the chance of another 9/11? It just doesn't follow.
But now you've got enough bandwidth to send video from the television back to the remote. No more of that stupid "picture in picture" crap, you can browse other channels on your remote while watching something else on the big screen.
Depending on your state, you can file a motion under SLAPP laws to protect yourself. If you are innocent of libel/slander, it is not allowable for a larger entity to quash your speech by threatening an expensive legal action.
I like Opera's gestures too, but have you tried the RadialContext plugin for Firefox (just Google it, I won't bother to link)? It's a lot like gestures, and having used it for almost a year now, I find it totally natural.
I switch browsers a lot -- I use Opera's gestures when it's around, Firefox with RadialContext when it's not, and CocoaGestures/Safari when I'm at home using my Mac Mini. It's kind of like being able to use both Emacs and vi comfortably -- totally possible to learn, and it makes you a lot more flexible.
(Speaking of Emacs and vi, have you ever tried running vi inside Emacs? Now there's a mindfuck.)
So, you continually test. Nobody said this was easy.
You're right, sorry. I left something out.
You can also do a P-set M-way N-fold cross validation, which is basically the procedure I outlined above, but repeated P times with P different data sets. The mechanics of computing the WLT table and the T-tests remain the same but obviously there are P times as many results. Now, you can simply start plugging in datasets until everybody is satisfied that the test is fair.
Thanks for pointing that out.
Typically what is done is to select a range of filters/learners that you want to evaluate. A test dataset is also selected (in this case, it would be an archive of spam and nonspam messages, correctly classified). An M-way N-fold cross validation is performed. What this means is that the data set is split into N parts, and N runs are conducted for each classifier, training using N-1 of the parts. The remaining part is used to test the learner. This is repeated, each time holding out a different part of the test set.
This ENTIRE procedure is repeated M times. This gives, ultimately, M*N results. Each column pair of results from a specific pair of learners has a T-test applied to it. This tells the statistical significance of variations in performance. Usually, a 5% or 1% threshold of significance is used.
Once that is completed, something called a WLT table is computed. Each time a learner defeats another learner on a given test, its W ("Win") counter is incremented. Likewise, when a learner loses, the L ("Loss") counter is incremented. When two learners tie (i.e., when the variation is not statistically significant), the T ("Tie") counter is incremented.
The overall "winner" of the comparison is the learner with the maximum value of W-L.
This sounds complicated and bloated, but it is, in fact, how machine learners are tested in academia. The cross validation method, along with checks for statistical significance, is critical to achieving a valid comparison. Simply running the tests once and saying "This filter got 98% correct, and this other filter got 95% correct -- therefore the first filter is better" is NOT sufficient.
Of course, and I addressed that in my post. Evolution happens via mutation. Selective breeding might take advantage of mutations when they occur, but it doesn't cause them. No ancient human could have ever bred a tomato into a corn plant.
Selecting the most desirable individuals and breeding them is not the same thing as genetic engineering. In selective breeding, the total gene pool remains fixed. You don't create new genes by breeding different plants together, you just create new combinations of existing genes. Every once in a while a mutation occurs, but we have no control over it. Genetic engineering is an entirely different thing -- we have learned how to modify, remove, and insert completely new genes into organisms, at will.
Given enough time, could you ever breed corn such that it became a tomato? It would never happen. Tomatos have genes that corn simply doesn't have. No amount of selection is going to magically turn corn into a tomato. But with our modern technology, we are able to take genes from the tomato and insert them into corn. So genetic engineering is really a vastly different thing than breeding domesticated crops.
just like dogs. do you love your dog? you're dog is a genetically modified wolf, warped by mankind into something wholly unnatural.
Once again, no. The genes of a domestic dog are indistinguishable from the genes of wolves. They simply happen to occur in an extremely unlikely combination, influenced by human breeders. It will never be possible, through breeding, to turn a dog into a sheep. But genetic engineering might one day make that possible.
What we are doing genetically to our food is absolutely nothing like selective breeding.
Totally wrong.
There is no "limit" to how much moisture the air can "hold." Water vapor and air are infinitely miscible, just like alcohol and water. The relative humidity can be far higher than 100% in certain situations.
Can you fax an RFID? I didn't think so.
I see. So, while my penis is inside a woman, she's part of my body. I can smack that bitch up!
I think your logic stinks.
Now that's a stupid way of looking at it. One line of code which executes billions of times is more important than a million lines which never execute.
And almost all of your routine interaction with the operating system is GNU-centric. If you switched the kernel to something else, it would still be fundamentally the same experience.
It's not GNU-centric, it's UNIX centric. And have you heard? GNU's not UNIX!
Even if you were using the best photovoltaic material available, you probably couldn't get more than 20 or 30 watts from a module small enough to carry around with you, and that's only when the sun is shining at its peak.
No matter what your solution, bear in mind that you will additionally need a charge controller and a battery to dump the energy into. You can't just plug a random power source directly into a device which expects clean DC at a specific voltage.
What logical thought occurs in the brain of an MDMA user that leads to the emotional state? There is none, it is the chemical which causes it. What you're saying is, "Effects have causes," well, duh. This has nothing to do with logic.