It sounds like this new feature (ahem) will allow independant compaines like VeriSign to certify software. Hopefully, they'll just let a developer get a certificate which will allow certification (really easy way to track the author of a virus).
Imagine what would happen to the Open Source community. Someone releases some app and signs it. Someone else patches it, and they need to sign the new one. Now every open-source contributor needs a certificate, and every build has its own signature. Ugh!
The upshot is that M$ is too lazy to implement a good sandbox (even JavaScript can thoroughly bring down Windows with looped window.open calls). If a program could be run with the equivalent of chroot, there would be no problem (except that windows is so convoluted that it is much more complicated to sandbox without breaking existing apps -- I imaging a virtual profile to run an untrusted app where all changes are app-specific and volatile)
It seems to me that the best solution is to release the specs for how the driver communicates w/ the device (all the specs, unlike how nVidia released partial specs). I don't think this information would compromise any IP. Then the Linux folks could write drivers for the device, maybe even better than the company made. Who knows, the new hacked-up drivers could provide some useful tricks that could be incorporated into the Windows drivers.
It seems to me that the major patents involved (6,049,846 and 6,067,592) were issued this year! Since SDRAM's been around for more than a few months, how can Rambus claim that they own the IP? If someone has a good explanation, I'd love to hear it.
If you have two counterrotating flywheels, then their angular momenta are in opposite directions, but since energy has no direction (it's a scalar), the energies just add. So you still have positive energy with counterrotating flywheels.
Slashdot Mirror
Imagine what would happen to the Open Source community. Someone releases some app and signs it. Someone else patches it, and they need to sign the new one. Now every open-source contributor needs a certificate, and every build has its own signature. Ugh!
The upshot is that M$ is too lazy to implement a good sandbox (even JavaScript can thoroughly bring down Windows with looped window.open calls). If a program could be run with the equivalent of chroot, there would be no problem (except that windows is so convoluted that it is much more complicated to sandbox without breaking existing apps -- I imaging a virtual profile to run an untrusted app where all changes are app-specific and volatile)
Working from memory (most likely wrong), 1 fundamental time unit = 10^-44 seconds = 2.78*10^-48 hours.
So 1 hour = 3.6*10^47, so 1Å*hour=3.3*10^25 hours. That's one hell of a fuel supply :)
P.S. Yes, I know it's really only 1-1/2 hours :(
It seems to me that the best solution is to release the specs for how the driver communicates w/ the device (all the specs, unlike how nVidia released partial specs). I don't think this information would compromise any IP. Then the Linux folks could write drivers for the device, maybe even better than the company made. Who knows, the new hacked-up drivers could provide some useful tricks that could be incorporated into the Windows drivers.
It seems to me that the major patents involved (6,049,846 and 6,067,592) were issued this year! Since SDRAM's been around for more than a few months, how can Rambus claim that they own the IP? If someone has a good explanation, I'd love to hear it.
If you have two counterrotating flywheels, then their angular momenta are in opposite directions, but since energy has no direction (it's a scalar), the energies just add. So you still have positive energy with counterrotating flywheels.