Am I the only that thinks that perhaps the programmers of the world can spend their energies doing something a little...um...i dunno...um...productive?
the additional noise energy UWB adds is very very minute in a single frequency range (UWB has incredibly low power spectral density by design; the energy of the signal is spread over a huge spectrum)
thats right, but the fear of UWB opponents is that when you have lots of UWB devices around (once it becomes mainstream), the noise may sum up to a level that it can jam the narrowband devices.
A nifty new processor feature won't be noticed in the computing world because it won't get used anyway,
this reminds me of the SIMD stuff (it's also a way of "parallel computing"). BTW, has the compiler support for automatic SIMD generation matured in the meantime?
Last time I checked, if you wanted to use MMX/SSE/etc. on x86, you had to hand-code your stuff in assembler (separately for PIII, PIV, AMD etc.), or buy an expensive Intel compiler.
In accepting the settlement Burst dropped the claim of copyright infringement.
I thought the case was about patent, not copyright infringement?
Not to be nit-picking here, but isn't it a big difference if you're accused of copyright infringement because you stole s.o. else's video streaming code, or if you're accused of patent infringement because some company holds a patent granting them a monopoly on streaming videos over The Internet(TM)?
But if someone's done gone and reverse-engineered stuff to the point of being able to write a substitute BIOS, doesn't that render hiding low-level hardware details kinda moot?
Theoretically, yes. But reverse-engineering is not as good as having access to specs. (and that's why RMS and others are asking to disclose them in the first place).
BTW, having people reverse engineer their hardware is not something they can influence, but they may choose to keep the specs secret (or offering access only under NDA).
Does Anyone know what Intel gets out of not opening it up? Are there any IP issues?
Due to the low-level nature of a BIOS, they would expose lots of hardware details. It's the same reason why many hardware manufacturers are reluctant to release open source drivers (or to provide OSS community with hardware specs).
How would you know if you did generate one? I mean can you tell by looking at a stream of numbers whether the output is "truly" random?
There are statistical tests (the chi-square test, for instance). They can tell how "random" a sequence of numbers is, and AFAIK they are also part of steganography research (you use them to tell if some image file, for instance, is "only" an image or if some hidden message has been encoded on top of it).
It depends on your requirements. We're using RTAI for a few applications and it gives us some 10 or 20 microsecond maximum interrupt jitter on a 100MHz PowerPC. For a lot of applications, this is OK. Why stick to a commercial RTOS if Linux solves the problem as well (at least if your RT requirements are not as tight).
Business wants it. Government wants what business wants.
That's simply not true. Business not only consists of big corporations, but small / medium enterprises as well. Over here in Europe, the majority of jobs are in small / medium enterprises (AFAIK). That's why politicians at least pretend to represent their interests.
After six years of that crap, there has to be something better out there.
Yes, there are not only big corporations out there. I guess that small / medium sized enterprises simply don't have the resources for such retarded games.
Re:My contrarian view of the GPL license
on
Why I Love The GPL
·
· Score: 1
Furthermore, after reviewing this GPL our lawyers advised us that any products compiled with GPL'ed tools - such as gcc - would also have to its source code released.
That's plain FUD and not true. You should consider hiring better lawyers.
I think you meant "unroll" it instead of parallelize (compilers don't split things up into threads as a general rule, although that's starting to change, in a way, with the 64 bit architectures)
FYI, parallelize can also mean something SIMD-like (SSE, MMX, AltiVec, you name it). If I'm not smoking crack, the previous poster's code example would also be broken if the copy operation was implemented using SIMD.
Slashdot Mirror
In some archaic societies, they even used female genital mutilation for that reason.
Like... posting on Slashdot?
thats right, but the fear of UWB opponents is that when you have lots of UWB devices around (once it becomes mainstream), the noise may sum up to a level that it can jam the narrowband devices.
But it can crash much faster than slow, buggy code.
this reminds me of the SIMD stuff (it's also a way of "parallel computing"). BTW, has the compiler support for automatic SIMD generation matured in the meantime?
Last time I checked, if you wanted to use MMX/SSE/etc. on x86, you had to hand-code your stuff in assembler (separately for PIII, PIV, AMD etc.), or buy an expensive Intel compiler.
They were not sued for hypocrisy, but because some other company has a patent on streaming video over networks. Is this a good thing?
I thought the case was about patent, not copyright infringement?
Not to be nit-picking here, but isn't it a big difference if you're accused of copyright infringement because you stole s.o. else's video streaming code, or if you're accused of patent infringement because some company holds a patent granting them a monopoly on streaming videos over The Internet(TM)?
Double standards, anyone?
You find it in the contiki CVS.
If law was software, it'd be the most inefficient, resource-wasting, bug-ridden piece of code ever.
Theoretically, yes. But reverse-engineering is not as good as having access to specs. (and that's why RMS and others are asking to disclose them in the first place).
BTW, having people reverse engineer their hardware is not something they can influence, but they may choose to keep the specs secret (or offering access only under NDA).
Due to the low-level nature of a BIOS, they would expose lots of hardware details. It's the same reason why many hardware manufacturers are reluctant to release open source drivers (or to provide OSS community with hardware specs).
Over here in germany, big governmental SW projects ("Toll Collect", "Hartz IV" and others come to mind) are infamous because they always fail.
But perhaps this is also due to shady call-for-bids practice.
There are statistical tests (the chi-square test, for instance). They can tell how "random" a sequence of numbers is, and AFAIK they are also part of steganography research (you use them to tell if some image file, for instance, is "only" an image or if some hidden message has been encoded on top of it).
> I'm talking about C and your &b declaration is not a valid C statement! Look at the parent - it talks about C!!
In C, it's even simpler.
#include <stdio.h>
#define b a
int main()
{
int a = 1234;
if (&a != &b) {
puts("true\n");
} else {
puts("false\n");
}
return 0;
}
&a != &b will always return TRUE!!!!!
./test
patrick@guantanamo:~/source$ cat >test.cpp
#include <iostream>
int main()
{
int a = 1234;
int &b = a;
if (&a != &b) {
std::cout << "true" << std::endl;
} else {
std::cout << "false" << std::endl;
}
return 0;
}
patrick@guantanamo:~/source$ g++ -o test test.cpp;
false
patrick@guantanamo:~/source$
It depends on your requirements. We're using RTAI for a few applications and it gives us some 10 or 20 microsecond maximum interrupt jitter on a 100MHz PowerPC. For a lot of applications, this is OK. Why stick to a commercial RTOS if Linux solves the problem as well (at least if your RT requirements are not as tight).
Patent applications must be technical, novel and non-obvious. For example, you can't patent "one-click shopping" or "progress bars".
Oh, wait...
That's simply not true. Business not only consists of big corporations, but small / medium enterprises as well. Over here in Europe, the majority of jobs are in small / medium enterprises (AFAIK). That's why politicians at least pretend to represent their interests.
Connect directly to the internet. Never use a NAT router, they're way too expensive.
Always use Internet Explorer. Firefox is too usable and has too many features anyway.
Run all obscure *.exe's that you get from spam email or file sharing.
Yes, there are not only big corporations out there. I guess that small / medium sized enterprises simply don't have the resources for such retarded games.
Why didn't he use JTAG?
That's plain FUD and not true. You should consider hiring better lawyers.
FYI, parallelize can also mean something SIMD-like (SSE, MMX, AltiVec, you name it). If I'm not smoking crack, the previous poster's code example would also be broken if the copy operation was implemented using SIMD.