The latter. They complained that Chrome does not let them hook into the networking component, so a similar add-on is impossible for Chrome. It may rewrite links, too, but that would not protect against external/manually typed in URLs or requests made via Javascript.
Excellent! What is the documentation for learning how to do this? Or, as a backup, where in the code for HTTPS Everywhere is the relevant piece of code?
Does it parse the webpage you are on and rewrite every link to use HTTPS or, better, does it intercept every request Firefox makes and rewrite that before it is sent?
The reason I'm interested is that I want to create an extension that does rewrites in the latter way described, but don't know how to do it.
Interesting! Are there internal (i.e., living inside the computer case) sound cards available for purchase that are as good as the one you are talking about? What would you recommend if I were to go out and buy a sound card today for recording use (say 4 or 8 channels) - internal or external? Any particular models or types you would recommend or warn against?
I thought people fly in rockets and visit space stations and the moon because it's cool. I don't care if no scientific progress comes out of it - I like space travel because it's awesome. Similarily, I'm not attracted to science, mathematics or technology for their practical uses, but because it's fun understanding how the world works, being able to calculate things and think up and admire cool (preferably huge) machines.
Good point, one should indeed try such though experiments to scrutinise one's own behaviour.
That said, a bias may be justified. If a trusted friend wants to borrow money, I let him, but if an infamous thief wants the same, he gets nothing from me.
Let's think of how they might have designed the algorithm for this. In the videos, it looks like it is treating one word at a time, so let's consider that scenario here as well. I would define the problem as assigning to every possible word the probability that it is the word the user intended. I would use Bayes' theorem to achieve this.
First, assume a prior probability distribution over all words. Words not in our dictionary, and words of the wrong length, we give probability zero. The remaining words can be assigned equal probability or, better, a probability proportional to their frequency in the language. If you want to be fancy, you could have more sophisticated models that knows which words are likely to come after others and such things.
Second, for each candidate word, what is the probability that the user would tap the screen as they just did? A model for this could be that the location of each tap is drawn from a Gaussian probability distribution centred at the intended letter with a known standard deviation and that each tap's deviation from its target is independent of the others'.
Finally, Bayes' theorem states that the posterior probability (the one we want to calculate) of each word is the word's prior probability (from step one) times its likelihood (the probability of step two).
To implement the arbitrary position, orientation and size of the keyboard, we redefine the problem from finding the probability of each candidate word to finding the probability of each tuple (intended word, keyboard position, keyboard orientation, keyboard size). Make it simple; have each element of this tuple to be independent of the other and use flat distributions for all keyboard parameters. To choose the most likely word, you could either pick the word of the most likely tuple or, more correctly, for each candidate word, integrate over all possible keyboard parameters (weighted with their prior probabilities) to get the probability of that word. Likewise, you could introduce the standard deviation of the taps as another element of the tuple, with its own prior distribution.
I suspect this method is a bit to heavy on computation cost and power consumption, so if you cannot find a clever way to do it fast, you might have to cut corners in the rigor (or do something completely different).
Cars have brakes. "Car breaks" means it stops working because of mechanical or electrical failure.
I honestly thought he was talking about car crashes and even though that was a strange way of saying it, I convinced myselft that is was physically sound.
I talked to a guy in Saint Louis once who was a genetic engineer for Monsanto. He didn't believe in evolution.
I don't think it's obvious that he would. I'm sure he believes that traits can be inherited and that by selecting who gets to reproduce, you can steer the new generations into having certain qualities, like breeding dogs to have long ears or whatever you fancy. Believing in evolution, on the other hand, would be to hold the position that the current plants and animals are the result of such a process, where the selection has been carried out by naturally occurring circumstances. Embracing evolution implies embracing genetics, but not the other way around.
I could give more details, but I'm afraid I've already said too much - because I certainly do not want to be identified by the company for fear of some kind of retaliation, either above or below the table
I understand your situation and I'm thankful that you told us what you did. Every little fragment of information is interesting for an outsider only has vague ideas about how people do it.
It's not a textbook; it assumes you basically know the math
That applies to every math book out there.
No, there is at least one mathematics book for which the statement does not hold. I don't have a constructive proof for this my claim, though, so I can't give you an example.
The math geek in me says that that's an engineering problem with the G and B channels on existing displays, since three independent measurements should mean that you only need three independent signal sources.
I see what you mean, but I'm not sure it's correct. Here's my speculation.;-) Please teach me if I'm wrong.
Let's view the spectrum of a ray of light as a mapping of each frequency to an intensity, i.e., a function of type (0,inf)->(0,inf). This is similar to a vector space (of infinite dimension), but it's not quite one, because we cannot have negative intensities. The three types of cones in our eyes are like three vectors in this "almost vector space", and by looking into the ray with the remaining eye we project the spectrum onto each of these three vectors, in total projecting it onto a three-dimensional subspace. Like the cones, the R, G and B ligth sources of a screen spans a three-dimensional subspace of the full spectrum space, though again, since there are no negative intensities, they don't actually span a full vector space.
Now, the problem is that when you go from a triple of signals in the screen to the spectrum they represent and finally to the triple of signals in the eye that this spectrum gets projected onto, there might be combinations of cone signals that you never reach. Indeed, this seems to be the case, judging from a glance at this colour space image. If we were dealing with true vector spaces, they would almost surely map nicely to each other, as you say.
All this migh also have the interesting consequence that there are combinations of signals from our cones that are not generated by any spectrum! I wonder what would happen if you induced those signals directly in the eye or nerve, without using light. Would we see new colours?
Slashdot Mirror
I think you got this CWB thing completely upside down.
The latter. They complained that Chrome does not let them hook into the networking component, so a similar add-on is impossible for Chrome. It may rewrite links, too, but that would not protect against external/manually typed in URLs or requests made via Javascript.
Excellent! What is the documentation for learning how to do this? Or, as a backup, where in the code for HTTPS Everywhere is the relevant piece of code?
Does it parse the webpage you are on and rewrite every link to use HTTPS or, better, does it intercept every request Firefox makes and rewrite that before it is sent?
The reason I'm interested is that I want to create an extension that does rewrites in the latter way described, but don't know how to do it.
Interesting! Are there internal (i.e., living inside the computer case) sound cards available for purchase that are as good as the one you are talking about? What would you recommend if I were to go out and buy a sound card today for recording use (say 4 or 8 channels) - internal or external? Any particular models or types you would recommend or warn against?
I thought people fly in rockets and visit space stations and the moon because it's cool. I don't care if no scientific progress comes out of it - I like space travel because it's awesome. Similarily, I'm not attracted to science, mathematics or technology for their practical uses, but because it's fun understanding how the world works, being able to calculate things and think up and admire cool (preferably huge) machines.
You'd be surprised what a trained and knowledgeable analyst can derive from just a few seemingly unconnected bits of information.
So, don't leave us in suspense then. Tell us! What did they find?
You just put a robots.txt file outside your door.
75 000 USD/year != 75 000 USD
Also, what is that uninformative picture of coins in a hand doing there? It does not add anything! This is just as bad as a newspaper article!
Good point, one should indeed try such though experiments to scrutinise one's own behaviour.
That said, a bias may be justified. If a trusted friend wants to borrow money, I let him, but if an infamous thief wants the same, he gets nothing from me.
I convinced myselft that is was physically sound.
I wasn't referring to the sound energy a car emits when it crashes, but maybe I should have.
Let's think of how they might have designed the algorithm for this. In the videos, it looks like it is treating one word at a time, so let's consider that scenario here as well. I would define the problem as assigning to every possible word the probability that it is the word the user intended. I would use Bayes' theorem to achieve this.
First, assume a prior probability distribution over all words. Words not in our dictionary, and words of the wrong length, we give probability zero. The remaining words can be assigned equal probability or, better, a probability proportional to their frequency in the language. If you want to be fancy, you could have more sophisticated models that knows which words are likely to come after others and such things.
Second, for each candidate word, what is the probability that the user would tap the screen as they just did? A model for this could be that the location of each tap is drawn from a Gaussian probability distribution centred at the intended letter with a known standard deviation and that each tap's deviation from its target is independent of the others'.
Finally, Bayes' theorem states that the posterior probability (the one we want to calculate) of each word is the word's prior probability (from step one) times its likelihood (the probability of step two).
To implement the arbitrary position, orientation and size of the keyboard, we redefine the problem from finding the probability of each candidate word to finding the probability of each tuple (intended word, keyboard position, keyboard orientation, keyboard size). Make it simple; have each element of this tuple to be independent of the other and use flat distributions for all keyboard parameters. To choose the most likely word, you could either pick the word of the most likely tuple or, more correctly, for each candidate word, integrate over all possible keyboard parameters (weighted with their prior probabilities) to get the probability of that word. Likewise, you could introduce the standard deviation of the taps as another element of the tuple, with its own prior distribution.
I suspect this method is a bit to heavy on computation cost and power consumption, so if you cannot find a clever way to do it fast, you might have to cut corners in the rigor (or do something completely different).
(Can I come work for them now?) :-)
Cars have brakes. "Car breaks" means it stops working because of mechanical or electrical failure.
I honestly thought he was talking about car crashes and even though that was a strange way of saying it, I convinced myselft that is was physically sound.
Surely, they have a web page, don't they?
I talked to a guy in Saint Louis once who was a genetic engineer for Monsanto. He didn't believe in evolution.
I don't think it's obvious that he would. I'm sure he believes that traits can be inherited and that by selecting who gets to reproduce, you can steer the new generations into having certain qualities, like breeding dogs to have long ears or whatever you fancy. Believing in evolution, on the other hand, would be to hold the position that the current plants and animals are the result of such a process, where the selection has been carried out by naturally occurring circumstances. Embracing evolution implies embracing genetics, but not the other way around.
I could give more details, but I'm afraid I've already said too much - because I certainly do not want to be identified by the company for fear of some kind of retaliation, either above or below the table
I understand your situation and I'm thankful that you told us what you did. Every little fragment of information is interesting for an outsider only has vague ideas about how people do it.
It's not a textbook; it assumes you basically know the math
That applies to every math book out there.
No, there is at least one mathematics book for which the statement does not hold. I don't have a constructive proof for this my claim, though, so I can't give you an example.
Key-value storage? That sounds like the ordinary file system to me.
The math geek in me says that that's an engineering problem with the G and B channels on existing displays, since three independent measurements should mean that you only need three independent signal sources.
I see what you mean, but I'm not sure it's correct. Here's my speculation. ;-) Please teach me if I'm wrong.
Let's view the spectrum of a ray of light as a mapping of each frequency to an intensity, i.e., a function of type (0,inf)->(0,inf). This is similar to a vector space (of infinite dimension), but it's not quite one, because we cannot have negative intensities. The three types of cones in our eyes are like three vectors in this "almost vector space", and by looking into the ray with the remaining eye we project the spectrum onto each of these three vectors, in total projecting it onto a three-dimensional subspace. Like the cones, the R, G and B ligth sources of a screen spans a three-dimensional subspace of the full spectrum space, though again, since there are no negative intensities, they don't actually span a full vector space.
Now, the problem is that when you go from a triple of signals in the screen to the spectrum they represent and finally to the triple of signals in the eye that this spectrum gets projected onto, there might be combinations of cone signals that you never reach. Indeed, this seems to be the case, judging from a glance at this colour space image. If we were dealing with true vector spaces, they would almost surely map nicely to each other, as you say.
All this migh also have the interesting consequence that there are combinations of signals from our cones that are not generated by any spectrum! I wonder what would happen if you induced those signals directly in the eye or nerve, without using light. Would we see new colours?
Their major search thing is as closed as they promised it wouldn't be (though no-one remembers that any more).
I didn't know they had promised that. Do you have a link?
Mind if I come and dump some toxic waste in your back garden then? ;-)
Any time. The entrance gate is open.
I don't care about the environment and I don't care about fraud, just stop putting "gate" at the end of everything!
Hmm.. government trying to dictate to the elderly what is moral in society.
It kind of does with every law created, when you think of it.
People falsely accused of pedophilia would beg to differ.
The rightly accused as well, I'd guess.
(Will our posts here be interpreted as a rise in pedophilia? Are we both now guilty by association for posting them?)
black holes were predicted to exist in the 18th Century(!)
Would you provide us with a link, please?
Most roads are already quite curvy in Europe and I'm pretty sure new roads are constructed in the same manner to encourage lower driving speeds.
Hey, it's the curvy roads that encourage "inspired driving"! Straights are only good for "steering lock" racing, which isn't much fun.