If they sell you 6Mb/s, but your connection would technically support 16Mb/s, how else should they make sure that you don't use the other 10Mb/s which they might have reserved for other customers paying for them?
They are oversold and they are throttling and NOT DELIVERING WHAT WE PAID FOR.
Maybe. But then the problem is not the throttling, but the overselling.
But to give you just what you paid for, in general the ISP will have to throttle. If it doesn't, you'll instead use all the bandwidth you can technically get, and others who have paid for that bandwidth might suffer.
Ah, I see it. You're quite right, and I see the point. Using pure set cardinality leaves the probability undefined, but this gives a workaround. I'd have some difficult reading to prove whether the original assertion is correct under measure theory, but I suspect it is.
And here's the poison pill to demonstrate how far I missed the boat: the same arguments I used apply to something as simple as picking a number x from [0,2] and computing the probability that x < 1. Obviously that's 0.5, but the technique I outlined would claim it's undefined.
I'm so glad I don't claim to make my living as a mathematician!
Actually it is undefined, until you've decided on a probability distribution. In the interval, you'll usually assume an uniform distribution, which gives your result 0.5, but that's not mandatory. It might be that the number is the square of an uniformly distributed random variable from the interval [0,sqrt(2)]. In which case, the probability would be slightly above 0.7.
I think you're here confusing the multiverses of string theory with those of the many worlds interpretation of quantum mechanics. These are completely orthogonal concepts (it's an unfortunate fact that the term "multiverse" is used in both contexts).
I think it's just another alibi to get government funding. Multiverse collision can be seen? Are you kidding me? If it is true then we must have been collided into pieces!
If I understand correctly, this is exactly what some string theorists think the big bang was.
I don't want to go there... I just strongly object to the notion that it is acceptable to prohibit or limit computers from performing what parallels, mimics, or otherwise effectively amounts to mental steps that can also be taken by a human being (albeit perhaps just not as conveniently or quickly) on nothing more than the basis that those very steps are perceived as some sort of rights violation (while at the same time not perceived as such when they are done by human beings).
For the legality it doesn't matter that it's computers doing it. It would be equally illegal if Facebook employed a few thousand employees to look at the images and tag the people there. The only difference is that Facebook wouldn't do that, because it would simply cost too much. The facial recognition software makes the task affordable, and thus moves the problem from a purely academic question to a practical one. But the actual problem isn't that it's a computer doing it. The problem is that it is done at all.
It seems to be, though, that the hardcore Linux base obsesses over customization and work. That's great. But apparently, "customization" means that you have to edit simple things in obscure config files deep in system directions, and "work" means that it has to look like a desktop from 1991.
About customization: The ideal system would be a system which works exactly the way I like it, out of the box. However, unfortunately this most likely won't happen, ever. Therefore it is important that I can customize the system, so that I'm not stuck more than necessary with decisions made by others. I don't care much if that customization is by configuration file or by GUI options, however I care that it's not overly hard to do. But there's one advantage of text files over GUI options: Virtually all text config files allow for comments, thus when you add/change a setting, you can, right at the point where you put the setting, add the information about *why* you put it. I've yet to find a GUI which allows that. Moreover, you can search for it, so if you ever wonder "how did I manage to enable feature X on this computer" (e.g. because you want to enable it on another computer as well), if you added a comment, you can easily grep for that comment and then see the option settings.
About "work": I care little about what it looks like. I care about the look mostly as far as it helps or distracts me from what I want to do. After all, I don't use the computer for the desktop environment. The desktop environment is a tool, not an end. If I want to put in a screw, I care whether the screwdriver fits the screw and whether it is easy to use. If it also looks good that's a bonus, but I'd always choose the screwdriver that works but is ugly over the screwdriver that's pretty but barely usable.
Shouldn't a simple reformat (i.e. rewriting the file system structures) already take care of it? OK, to be completely sure, also rewrite the MBR (there might be a boot sector virus on the disk as well). But I don't see the need to overwrite everything on the disk. It's not private or secret data you want to hide from others. You don't need to erase every trace of it; you just need to make sure the computer won't execute it.
Also, I'm sure that as soon as that algorithm is available, one use for it will be by men impersonating women, in order to see whether their texts pass the test, and adapt accordingly.
Well, in theory one could possibly craft a picture which uses a bug in the camera so that if it is photographed in the right way, it will create an executable to be executed when the camera gets connected.
In practice, of course, even if a camera turned out to have such a bug, the probability of the picture being taken exactly at the exact right distance with the exact right angle at the exactly correct light situation should be that low that winning the lottery six times in a row would seem like a sure bet in comparison.:-)
If you have FTL, you don't need any curvature. However with appropriate curvature, you don't need FTL. The non-FTL version of time travel is also known as closed timelike curve.
But the point here is, FTL enables time travel without any curvature. Note that in that case your world line is not timelike.
If two photons collide, it's not a decay. Most importantly, the photons don't need mass for that, since energy and momentum conservation can be fulfilled even with massless particles (since the photons have different momentum -- otherwise they wouldn't collide --, there's a frame of reference where the total momentum is zero; if in that frame of reference the energy is at least 2mc^2 (with m the electron mass), an electron-positron pair can be generated (however the probability of that happening is extremely low; I don't think you could actually observe it in experiment). A single particle needs rest mass to decay because otherwise it would be impossible to fulfil both energy and momentum conservation. For two colliding photons producing an electron-positron pair this is no problem as long as the energy is high enough.
Can you show me a calculation in relativistic quantum mechanics where a physical photon or other particle (i.e. not just a non-observable intermediate result or single term with no isolated physical meaning, like virtual particles) goes faster than light? From my understanding, the fact that spacelike separated field operators commute guarantees that this cannot happen. But then, I admit that quantum field theory isn't my field of expertise, so there might be something I don't know. However I won't believe it until I see it (in the form of a valid calculation showing a physical(!) particle going faster than light).
Slashdot Mirror
I didn't RTFA, but note that the word "chip" is older than the electronic chip. So I guess the thing has just the physical form of a chip.
If they sell you 6Mb/s, but your connection would technically support 16Mb/s, how else should they make sure that you don't use the other 10Mb/s which they might have reserved for other customers paying for them?
Maybe. But then the problem is not the throttling, but the overselling.
But to give you just what you paid for, in general the ISP will have to throttle. If it doesn't, you'll instead use all the bandwidth you can technically get, and others who have paid for that bandwidth might suffer.
They found DOS there? I didn't know Siemens S7 was running under ancient operating systems. :-)
Ah, I see it. You're quite right, and I see the point. Using pure set cardinality leaves the probability undefined, but this gives a workaround. I'd have some difficult reading to prove whether the original assertion is correct under measure theory, but I suspect it is.
And here's the poison pill to demonstrate how far I missed the boat: the same arguments I used apply to something as simple as picking a number x from [0,2] and computing the probability that x < 1. Obviously that's 0.5, but the technique I outlined would claim it's undefined.
I'm so glad I don't claim to make my living as a mathematician!
Actually it is undefined, until you've decided on a probability distribution. In the interval, you'll usually assume an uniform distribution, which gives your result 0.5, but that's not mandatory. It might be that the number is the square of an uniformly distributed random variable from the interval [0,sqrt(2)]. In which case, the probability would be slightly above 0.7.
The usual term is "hyperplane". More exactly, in an n-dimensional space a hyperplane is an (n-1)-dimensional subspace.
And yet others wasted their time posting comments on Slashdot about it ...
I think you're here confusing the multiverses of string theory with those of the many worlds interpretation of quantum mechanics. These are completely orthogonal concepts (it's an unfortunate fact that the term "multiverse" is used in both contexts).
Only if we can devise another test to show that the other universes experienced the big band at the same time as us.
I think some of them experienced a symphonic orchestra instead.
If I understand correctly, this is exactly what some string theorists think the big bang was.
in this universe at least.
COOLEST "first post" EVER.
In your universe, maybe.
For the legality it doesn't matter that it's computers doing it. It would be equally illegal if Facebook employed a few thousand employees to look at the images and tag the people there. The only difference is that Facebook wouldn't do that, because it would simply cost too much. The facial recognition software makes the task affordable, and thus moves the problem from a purely academic question to a practical one. But the actual problem isn't that it's a computer doing it. The problem is that it is done at all.
About customization: The ideal system would be a system which works exactly the way I like it, out of the box. However, unfortunately this most likely won't happen, ever. Therefore it is important that I can customize the system, so that I'm not stuck more than necessary with decisions made by others. I don't care much if that customization is by configuration file or by GUI options, however I care that it's not overly hard to do. But there's one advantage of text files over GUI options: Virtually all text config files allow for comments, thus when you add/change a setting, you can, right at the point where you put the setting, add the information about *why* you put it. I've yet to find a GUI which allows that. Moreover, you can search for it, so if you ever wonder "how did I manage to enable feature X on this computer" (e.g. because you want to enable it on another computer as well), if you added a comment, you can easily grep for that comment and then see the option settings.
About "work": I care little about what it looks like. I care about the look mostly as far as it helps or distracts me from what I want to do. After all, I don't use the computer for the desktop environment. The desktop environment is a tool, not an end. If I want to put in a screw, I care whether the screwdriver fits the screw and whether it is easy to use. If it also looks good that's a bonus, but I'd always choose the screwdriver that works but is ugly over the screwdriver that's pretty but barely usable.
Shouldn't a simple reformat (i.e. rewriting the file system structures) already take care of it? OK, to be completely sure, also rewrite the MBR (there might be a boot sector virus on the disk as well). But I don't see the need to overwrite everything on the disk. It's not private or secret data you want to hide from others. You don't need to erase every trace of it; you just need to make sure the computer won't execute it.
Also, I'm sure that as soon as that algorithm is available, one use for it will be by men impersonating women, in order to see whether their texts pass the test, and adapt accordingly.
That solves the image refocusing part. However how would you detect the distance at which the eyes focus?
There's a third category where Windows is the clear winner: Even the most obscure hardware device tends to have Windows drivers.
Well, in theory one could possibly craft a picture which uses a bug in the camera so that if it is photographed in the right way, it will create an executable to be executed when the camera gets connected.
In practice, of course, even if a camera turned out to have such a bug, the probability of the picture being taken exactly at the exact right distance with the exact right angle at the exactly correct light situation should be that low that winning the lottery six times in a row would seem like a sure bet in comparison. :-)
What about Mercurial?
I can imagine some countries would rather do all the snooping behind peoples back.
Like all the western nations - google ECHELON
Not all western countries do it (only) behind people's back ...
https://secure.wikimedia.org/wikipedia/en/wiki/Telecommunications_data_retention
If you have FTL, you don't need any curvature. However with appropriate curvature, you don't need FTL. The non-FTL version of time travel is also known as closed timelike curve.
But the point here is, FTL enables time travel without any curvature. Note that in that case your world line is not timelike.
But if you look closer, you'll find that what had been intended to be a Lisp application was mostly hacked together in Perl.
Does it also say you should not be afraid of a Whoosh?
If two photons collide, it's not a decay. Most importantly, the photons don't need mass for that, since energy and momentum conservation can be fulfilled even with massless particles (since the photons have different momentum -- otherwise they wouldn't collide --, there's a frame of reference where the total momentum is zero; if in that frame of reference the energy is at least 2mc^2 (with m the electron mass), an electron-positron pair can be generated (however the probability of that happening is extremely low; I don't think you could actually observe it in experiment). A single particle needs rest mass to decay because otherwise it would be impossible to fulfil both energy and momentum conservation. For two colliding photons producing an electron-positron pair this is no problem as long as the energy is high enough.
Can you show me a calculation in relativistic quantum mechanics where a physical photon or other particle (i.e. not just a non-observable intermediate result or single term with no isolated physical meaning, like virtual particles) goes faster than light? From my understanding, the fact that spacelike separated field operators commute guarantees that this cannot happen. But then, I admit that quantum field theory isn't my field of expertise, so there might be something I don't know. However I won't believe it until I see it (in the form of a valid calculation showing a physical(!) particle going faster than light).