wifi can share spectrum to to tight regulations on broadcast power and antenna design, and that the major protocol running over it, 802.11x has been designed to cooperate more-or-less with itself, and to avoid low levels of narrowband interference (through DSSS). If the 2.4 GHz band were anywhere near capacity with a heterogeneous collection of devices/protocols, it would completely fail. Even so, it is fairly susceptible to outside interference.
But yes, for the most part whoever broadcast the loudest wins, but why not
Because that is a retarded way to decide things. Plus, it doesn't even work that well--nobody gets heard, because it takes less power to disrupt communication than to use it.
And no two stations reallly want to be broadcasting on the same frequency as neither of them would get heard.
Right. Ideally, I would broadcast on my frequency, and jam your frequency so nobody listens, and I rake in the big bucks. Or I jam your station because I don't like you and you are a giant corporation, or political talk radio, or whatever.
whoever has the money to broadcast loud has the money to buy spectrum from the government
You are off by several orders of magnitude. The FCC is talking about 10 billion dollars here. The cost to build a jammer that will disrupt a radio or TV station in a major metropolitan area is measured in thousands of dollars.
Seriously, this is economics 101. A free-for-all is not the same thing as a free market.
The (in)decency regulation is an entirely separate issue that has no bearing on whether the FCC should license broadcast spectrum.
Oh rubbish. The template code will be the same as, or smaller than, the non-template code. Unless your compiler is truly awful.
That entirely depends on how you use templates. A single instantiation of a template will likely be smaller than the equivalent generic code, especially in a complex class where you only have to instantiate the members you use. However, if you allow the same code to be instantiated several times, it will quickly blow up your executable size. Any way you avoid this, you quickly lose the benefit of templates (i.e., you make the generic version and lose compile time type-checking). Have you seen the size of executables that heavily use STL? They can be huge.
What 'leaner thing' can you do in C but not in C++?
Since most C programs are valid C++ programs, and will compile to a comparable size, almost nothing. However, as soon as you actually *use* non-trivial C++ features, such as the standard library, you are lost on small platforms. Any programming language you can't like with the most basic element of the languages standard library without taking up 90% of your target RAM is... useless (for that target).
Actually, "corporate personhood" only protects the owners (i.e., shareholders), not employees. Employees of a non-corporate company (i.e., a partnership) are no more or less liable for their actions that those of a large corporation. In principle, employees are responsible for their actions, and management is responsible for actions of their underlings, though in practice it is hard to convict people, this has nothing to do per se with their status as a corporation. The difference is, in a partnership, if an employee screws up while performing duties for the company, all partners have potentially unlimited financial liability. In a corporation, liability to the owners only extends to the assets held by the corporation.
There are lots of other ways in which corporations are treated as people, but most of it comes down to "they are non-person entities which can own property" -- this is the root of their ability to be slandered or libeled, their ability to be a party to a lawsuit, and so forth.
Semiconductor fabrication is already a commodity business. Outside of Intel, there are essentially no vertically integrated semiconductor manufacturers left. Lots of people can design silicon, incorporate all or part of the UltraSparc core into it, and send it off to fab. For instance, NVidia could base their GPUs off of the Niagara arch.
It isn't something that every kid and his grandma do in their garage, but lots of companies (even relatively small ones) could actually use this. Whether they will, or it will turn out to be useful, is hard to say.
I think what he is saying, which you have either missed or ignored completely is that once you have a PhD you should be a (possibly junior) partner, not a student with a mentor. A PhD is a research degree, and the purpose is to train you to learn things that nobody knows. Once reach that level, the field matters less--as long as you have solid knowledge of the fundamentals (i.e., at a bachelors or masters level). Hence the car analogy.
That is true, but that isn't what is happening here, according to the article.
According to the article, this happens when you click on a mailto: link with escaped null bytes in it, and instead of launching the registered mail client (i.e., outlook ), a command specified in the URI (calc.exe) is executed. This seems to work regardless of which URI scheme is used, and regardless of what the associated handler is. Sounds like a pretty cut-and-dry windows bug to me.
No, we have evidence, though. Other scientists have done similar (and better) experiments, and they have all shown that entanglement cannot be used to non-locally transmit information. Entanglement correlations only show up "in coincidence" -- that is, when comparing the results of the two sides after the fact.
Lots of people get sick or die from food poisoning. Nobody to my knowledge has died from GM foods, irradiated foods, or growth hormone. There are no scientific studies that show negative effects from the first two.
I find the occasionally excessive enthusiasm for pasteurization annoying, but it isn't like there is no reason for it.
To be fair, the things that are included in the "advanced" kit are development boards and a fancy carrying case, along with some duplicate components like memory cards and USB cables which you can probably pick up cheaper elsewhere.
The base kit looks like it has everything you would need or want for a user or software developer. The advanced kit is for hardware hackers.
The consumer version is going to be $450 for the base model, though it has wifi and a bunch of other neato hardware added.
The real reason it looks expensive is that it comes unlocked. If you could get a $150 rebate on a cell phone contract if you bought your own handset, suddenly it would look a lot more favorable compared to other smart phones.
Says you. If my granchildren don't have the possibility to live that long, I am going to be disappointed. I haven't dismissed the possiblity that I will still be around then, in some form or another.
At least on my dell (not sure how Macs handle this), the power button generates an ACPI event when the power button is pressed, not released. Thus, you could have a software alarm go off for 5 seconds while the would be thief is holding the power button down. Likewise, you can lift up the laptop and pop out the battery, but that probably sets of the alarm first.
For someone worried about a someone grabbing their laptop when they are getting a refill on their coffee, the motion sensor alarm should work pretty well. I was wishing the other day that my laptop had a tilt sensor so I could rig something like that up.
Washington DC has three electoral votes. Technically, Per the 23rd amendment, they have as many electors votes as their population would entitle them to if they were a state, not to exceed the least populous state.
In the 2000 election, one elector abstained to protest the lack of representation in congress.
Well, it is certainly now relatively easy for LANL employees to communicate via the unclassified network. Lots of non-classified research there is done in collaboration with people at universities and other labs, and that requires communication.
There is a fairly good argument that that work should not be done at LANL, but as long as it is, they need realtively accessable public communication. Another consideration is that for some non-classified research, the govt. also wants the ability to classify parts of the work if they deem it necessary. Much easier if it is done in a secure national lab.
Try looking at it from the other end. Look at people addicted to gambling, people addicted to video games, and people addicted to porn. Now, if those groups show differences beyond the thing to which they are addicted, you should treat them as at least somewhat different conditions.
But ive got an information loss problem of my own. I got this neat random number generator with a lcd display that generates a random number every second. Ever since i placed in inside a black box i dont know what number it displays. Anyone know how to find out what number it displays without opening the box?
Any time we try to look at such things at the microscopic level. the history of which random numbers were displayed is recorded somewhere (often many places), for instance in the motional states of the atoms that make up the box and lcd (aka heat). It is the same "paradox" as exists in classical mechanics vs. statistical mechanics -- individual interactions are always reversable via CPT symmetry. Only when you blur things enough you can't see the individual interactions do you get properties like temperature.
The difference is that a particle falling into a black hole and being emitted as hawking radiation appears to be fundamentally non-unitary (quantum mechanical speak for non-reversable / information losing). Many people find it odd, and unlikely that every process in the universe is unitary except around black holes. Hence attempts to find alternative theories.
No conclusive answer will be found until we come up with a unified theory of quantum gravity and experimentally verify it. However, examining the weird boundaries between the quantum world and GR will help us understand what we are looking for in a unified theory.
Yes. Also, myth has content-based commercial detection using blank frames, audio dynamic range, and scene cut rate. For most programs, a single button jumps an entire block of commercials regardless of the length. You can set myth to automatically cut commercials, and literally never see another commercial at any speed. It doesn't do very well catching very short segments such as the "moment of zen"/credits at the end of the daily show, so I don't do that, but the auto-skip is far better than tivo's FF overshoot compensation.
I think using the data stream in a commercial application/device was a violation of the TOS. The small number of pre-built mythtv systems for sale might constitute a violation, though I suspect that is a tiny amount.
What would be a problem for them is if TiVo or another commercial DVR, got an indivdual subscription to their service, and used that to do their client updates. This is a big part of their target market, so losing it would be a big deal. The suspicious part of me wonders if such a company did that deliberately to force zap2it to turn off their service, crippling their OSS competition.
Usually, when marginally responsible people talk about cyber-terrorism, they mean either using the internet to launch attacks against sensitive targets (power plants, air traffic control, etc.) with the intent of damage to property or person, or to disrupt communication and prevent effective response to a coincident attack (i.e., as one part of a larger terrorist attack).
The reason it is a red herring is that most of the first type of attacks have not currently been shown effective compared to "conventional" methods, while in some cases our response without jammed communications is bad enough the second type is ineffective.
You could make arguments that using the internet to steal money to fund terrorism is terrorism, but that is probably a stretch, assuming you want to avoid the definition "terrorism is acts/crimes committed by terrorists".
You can work around a great deal of limitations in a filesystem at the application level, whether by partitioning drives due to max FS size or files due to max file size, or nesting directories with too many entries to access quickly. That doesn't mean it isn't annoying or that a better solution is not required in the long run.
Sure, I can make my video camera split files, but now I have to take care to move the bunch of them together, or reassemble them when downloading onto my computer. Also, if I have a large movie file I want to transfer between computers that don't have filesize limits, I am screwed again. If I want to hold a disk image for virtualization, I have to hope my virtualization software supports splitting that.
We could squeeze a few more years out of FAT32 by making a standard for splitting large files by with a decorated filename (bigfile.mov;1, bigfile.mov;2, and bigfile.mov;3) in a way that could be automatically handled by the FS layer on your OS, but that seems like kind of a wasted effort in light of the performance issues and the maximum volume size.
I think the way it is supposed to work (no idea how well it does for the linux kernel) is that anybody can "join" the development tree, publish "their" kernel version, including whichever patches they want, without interfereing with each other. Everybody has merge privileges on their own tree, and nobody has merge privileges on someone else's. The only reason Linus is "special" is that when he says "this is version 2.6.22, no bloody -mm -ac or -aa" everybody else listens.
Ideally, such a system would blur the distiction between a "forked" and "not forked" state. Typically, in OSS devlopment when two developers have unreconcilable differences in the direction they think a project should go, it forks. One or both parties creates a new project, a new CVS/SVN/whatever tree, and the two projects diverge. In the git model, they both stay in the same scm universe, and either party (or a third party) can pick any improvements from the other "team". Eventually, they might merge back into one project, without even having to decide whose VC tree to use.
In theory, this could really cut down on a lot of political arguments. As I said, I haven't followed enough kernel development recently to see how well it works in practice.
Nuclear certainly has its share of problems, but even at the current state of the art it is probably better than coal and natural gas. I think that our current proven deposits of uranium could supply the entire US electricity demands for a couple of decades, and if U238 breeder reactors become economical, many times that long.
It is also something that we can begin redeploying soon, while R&D on fusion and renewables continues. Nuclear is currently the only technology which shows an ability to take over a significant fraction of base load generation in the near future.
Slashdot Mirror
Because that is a retarded way to decide things. Plus, it doesn't even work that well--nobody gets heard, because it takes less power to disrupt communication than to use it.
Right. Ideally, I would broadcast on my frequency, and jam your frequency so nobody listens, and I rake in the big bucks. Or I jam your station because I don't like you and you are a giant corporation, or political talk radio, or whatever.
You are off by several orders of magnitude. The FCC is talking about 10 billion dollars here. The cost to build a jammer that will disrupt a radio or TV station in a major metropolitan area is measured in thousands of dollars.
Seriously, this is economics 101. A free-for-all is not the same thing as a free market.
The (in)decency regulation is an entirely separate issue that has no bearing on whether the FCC should license broadcast spectrum.
That entirely depends on how you use templates. A single instantiation of a template will likely be smaller than the equivalent generic code, especially in a complex class where you only have to instantiate the members you use. However, if you allow the same code to be instantiated several times, it will quickly blow up your executable size. Any way you avoid this, you quickly lose the benefit of templates (i.e., you make the generic version and lose compile time type-checking). Have you seen the size of executables that heavily use STL? They can be huge.
Since most C programs are valid C++ programs, and will compile to a comparable size, almost nothing. However, as soon as you actually *use* non-trivial C++ features, such as the standard library, you are lost on small platforms. Any programming language you can't like with the most basic element of the languages standard library without taking up 90% of your target RAM is... useless (for that target).
Actually, "corporate personhood" only protects the owners (i.e., shareholders), not employees. Employees of a non-corporate company (i.e., a partnership) are no more or less liable for their actions that those of a large corporation. In principle, employees are responsible for their actions, and management is responsible for actions of their underlings, though in practice it is hard to convict people, this has nothing to do per se with their status as a corporation. The difference is, in a partnership, if an employee screws up while performing duties for the company, all partners have potentially unlimited financial liability. In a corporation, liability to the owners only extends to the assets held by the corporation.
There are lots of other ways in which corporations are treated as people, but most of it comes down to "they are non-person entities which can own property" -- this is the root of their ability to be slandered or libeled, their ability to be a party to a lawsuit, and so forth.
Semiconductor fabrication is already a commodity business. Outside of Intel, there are essentially no vertically integrated semiconductor manufacturers left. Lots of people can design silicon, incorporate all or part of the UltraSparc core into it, and send it off to fab. For instance, NVidia could base their GPUs off of the Niagara arch.
It isn't something that every kid and his grandma do in their garage, but lots of companies (even relatively small ones) could actually use this. Whether they will, or it will turn out to be useful, is hard to say.
I think what he is saying, which you have either missed or ignored completely is that once you have a PhD you should be a (possibly junior) partner, not a student with a mentor. A PhD is a research degree, and the purpose is to train you to learn things that nobody knows. Once reach that level, the field matters less--as long as you have solid knowledge of the fundamentals (i.e., at a bachelors or masters level). Hence the car analogy.
Luckily, you don't design cars, and Toyota engineers have considered this strategy and decided that it is not as good as their current system.
What gave you away was the NO EXCEPTIONS in all caps. That is the mark of a religious fanatic -- not someone who should be allowed near a CAD program.
That is true, but that isn't what is happening here, according to the article.
According to the article, this happens when you click on a mailto: link with escaped null bytes in it, and instead of launching the registered mail client (i.e., outlook ), a command specified in the URI (calc.exe) is executed. This seems to work regardless of which URI scheme is used, and regardless of what the associated handler is. Sounds like a pretty cut-and-dry windows bug to me.
No, we have evidence, though. Other scientists have done similar (and better) experiments, and they have all shown that entanglement cannot be used to non-locally transmit information. Entanglement correlations only show up "in coincidence" -- that is, when comparing the results of the two sides after the fact.
You mean you're not a real doctor?
Lots of people get sick or die from food poisoning. Nobody to my knowledge has died from GM foods, irradiated foods, or growth hormone. There are no scientific studies that show negative effects from the first two.
I find the occasionally excessive enthusiasm for pasteurization annoying, but it isn't like there is no reason for it.
To be fair, the things that are included in the "advanced" kit are development boards and a fancy carrying case, along with some duplicate components like memory cards and USB cables which you can probably pick up cheaper elsewhere.
The base kit looks like it has everything you would need or want for a user or software developer. The advanced kit is for hardware hackers.
The consumer version is going to be $450 for the base model, though it has wifi and a bunch of other neato hardware added.
The real reason it looks expensive is that it comes unlocked. If you could get a $150 rebate on a cell phone contract if you bought your own handset, suddenly it would look a lot more favorable compared to other smart phones.
Says you. If my granchildren don't have the possibility to live that long, I am going to be disappointed. I haven't dismissed the possiblity that I will still be around then, in some form or another.
At least on my dell (not sure how Macs handle this), the power button generates an ACPI event when the power button is pressed, not released. Thus, you could have a software alarm go off for 5 seconds while the would be thief is holding the power button down. Likewise, you can lift up the laptop and pop out the battery, but that probably sets of the alarm first.
For someone worried about a someone grabbing their laptop when they are getting a refill on their coffee, the motion sensor alarm should work pretty well. I was wishing the other day that my laptop had a tilt sensor so I could rig something like that up.
Washington DC has three electoral votes. Technically, Per the 23rd amendment, they have as many electors votes as their population would entitle them to if they were a state, not to exceed the least populous state.
In the 2000 election, one elector abstained to protest the lack of representation in congress.
Well, it is certainly now relatively easy for LANL employees to communicate via the unclassified network. Lots of non-classified research there is done in collaboration with people at universities and other labs, and that requires communication.
There is a fairly good argument that that work should not be done at LANL, but as long as it is, they need realtively accessable public communication. Another consideration is that for some non-classified research, the govt. also wants the ability to classify parts of the work if they deem it necessary. Much easier if it is done in a secure national lab.
Try looking at it from the other end. Look at people addicted to gambling, people addicted to video games, and people addicted to porn. Now, if those groups show differences beyond the thing to which they are addicted, you should treat them as at least somewhat different conditions.
Any time we try to look at such things at the microscopic level. the history of which random numbers were displayed is recorded somewhere (often many places), for instance in the motional states of the atoms that make up the box and lcd (aka heat). It is the same "paradox" as exists in classical mechanics vs. statistical mechanics -- individual interactions are always reversable via CPT symmetry. Only when you blur things enough you can't see the individual interactions do you get properties like temperature.
The difference is that a particle falling into a black hole and being emitted as hawking radiation appears to be fundamentally non-unitary (quantum mechanical speak for non-reversable / information losing). Many people find it odd, and unlikely that every process in the universe is unitary except around black holes. Hence attempts to find alternative theories.
No conclusive answer will be found until we come up with a unified theory of quantum gravity and experimentally verify it. However, examining the weird boundaries between the quantum world and GR will help us understand what we are looking for in a unified theory.
Yes. Also, myth has content-based commercial detection using blank frames, audio dynamic range, and scene cut rate. For most programs, a single button jumps an entire block of commercials regardless of the length. You can set myth to automatically cut commercials, and literally never see another commercial at any speed. It doesn't do very well catching very short segments such as the "moment of zen"/credits at the end of the daily show, so I don't do that, but the auto-skip is far better than tivo's FF overshoot compensation.
I think using the data stream in a commercial application/device was a violation of the TOS. The small number of pre-built mythtv systems for sale might constitute a violation, though I suspect that is a tiny amount.
What would be a problem for them is if TiVo or another commercial DVR, got an indivdual subscription to their service, and used that to do their client updates. This is a big part of their target market, so losing it would be a big deal. The suspicious part of me wonders if such a company did that deliberately to force zap2it to turn off their service, crippling their OSS competition.
Usually, when marginally responsible people talk about cyber-terrorism, they mean either using the internet to launch attacks against sensitive targets (power plants, air traffic control, etc.) with the intent of damage to property or person, or to disrupt communication and prevent effective response to a coincident attack (i.e., as one part of a larger terrorist attack).
The reason it is a red herring is that most of the first type of attacks have not currently been shown effective compared to "conventional" methods, while in some cases our response without jammed communications is bad enough the second type is ineffective.
You could make arguments that using the internet to steal money to fund terrorism is terrorism, but that is probably a stretch, assuming you want to avoid the definition "terrorism is acts/crimes committed by terrorists".
You can work around a great deal of limitations in a filesystem at the application level, whether by partitioning drives due to max FS size or files due to max file size, or nesting directories with too many entries to access quickly. That doesn't mean it isn't annoying or that a better solution is not required in the long run.
Sure, I can make my video camera split files, but now I have to take care to move the bunch of them together, or reassemble them when downloading onto my computer. Also, if I have a large movie file I want to transfer between computers that don't have filesize limits, I am screwed again. If I want to hold a disk image for virtualization, I have to hope my virtualization software supports splitting that.
We could squeeze a few more years out of FAT32 by making a standard for splitting large files by with a decorated filename (bigfile.mov;1, bigfile.mov;2, and bigfile.mov;3) in a way that could be automatically handled by the FS layer on your OS, but that seems like kind of a wasted effort in light of the performance issues and the maximum volume size.
I think the way it is supposed to work (no idea how well it does for the linux kernel) is that anybody can "join" the development tree, publish "their" kernel version, including whichever patches they want, without interfereing with each other. Everybody has merge privileges on their own tree, and nobody has merge privileges on someone else's. The only reason Linus is "special" is that when he says "this is version 2.6.22, no bloody -mm -ac or -aa" everybody else listens.
Ideally, such a system would blur the distiction between a "forked" and "not forked" state. Typically, in OSS devlopment when two developers have unreconcilable differences in the direction they think a project should go, it forks. One or both parties creates a new project, a new CVS/SVN/whatever tree, and the two projects diverge. In the git model, they both stay in the same scm universe, and either party (or a third party) can pick any improvements from the other "team". Eventually, they might merge back into one project, without even having to decide whose VC tree to use.
In theory, this could really cut down on a lot of political arguments. As I said, I haven't followed enough kernel development recently to see how well it works in practice.
I think he was talking about the 4GB filesize limit, which is a serious problem for video recording.
That is the stupidest thing I ever heard. Unfortunately, that is often the way things *do* work, but it should not be.
Nuclear certainly has its share of problems, but even at the current state of the art it is probably better than coal and natural gas. I think that our current proven deposits of uranium could supply the entire US electricity demands for a couple of decades, and if U238 breeder reactors become economical, many times that long.
It is also something that we can begin redeploying soon, while R&D on fusion and renewables continues. Nuclear is currently the only technology which shows an ability to take over a significant fraction of base load generation in the near future.