> Crusoe, in optimizing execution, has the benefit of *knowing* what's happening at run time, as opposed to a normal compiler that has to guess about it at compile time
But this is not insurmountable. I know of at least one compiler that is available in the embedded market where you can optimize based on real execution results with an instrumented compile.
The clever thing I've seen is the talk about dynamically changeing the power drain (by changing speed) depending on how loaded the processor is. How does it know if its is running an idle loop or a boring iterative calculation?
Where TAOS approached the problem of binary compatibility between platforms with the Virtual Processor concept the Crusoe solution looks like the exact opposite.
The have accepted that x86 IS the market (at least volume wise) so have focused on that. I'm not supprised that they were cool on the idea of porting the CodeMorphing engine to PowerPC for the Mac. IMB/Motorola have a lot less problems with the power consumption of these chips as they are not tied to Wintel legacies.
Having said that there is no reason why you can't run native - and everybodies favorite OS seems to be the easiest to port.
> Closed source programs should probably never be allowed an setuid bit as closed source programmers tend to be sloppier and their source isn't open to review.
I have to take issue with that statement. I agree that Open source has benifits in public exposure, but please don't assume that all closed source programmers are sloppier just because the public (who 99% don't look at the code) don't see the code. Code Review is an important part of ny software development (open or closed).
Slashdot Mirror
> Crusoe, in optimizing execution, has the benefit of *knowing* what's happening at run time, as opposed to a normal compiler that has to guess about it at compile time
But this is not insurmountable. I know of at least one compiler that is available in the embedded market where you can optimize based on real execution results with an instrumented compile.
The clever thing I've seen is the talk about dynamically changeing the power drain (by changing speed) depending on how loaded the processor is. How does it know if its is running an idle loop or a boring iterative calculation?
Where TAOS approached the problem of binary compatibility between platforms with the Virtual Processor concept the Crusoe solution looks like the exact opposite.
The have accepted that x86 IS the market (at least volume wise) so have focused on that. I'm not supprised that they were cool on the idea of porting the CodeMorphing engine to PowerPC for the Mac. IMB/Motorola have a lot less problems with the power consumption of these chips as they are not tied to Wintel legacies.
Having said that there is no reason why you can't run native - and everybodies favorite OS seems to be the easiest to port.
> Closed source programs should probably never be allowed an setuid bit as closed source programmers tend to be sloppier and their source isn't open to review.
I have to take issue with that statement. I agree that Open source has benifits in public exposure, but please don't assume that all closed source programmers are sloppier just because the public (who 99% don't look at the code) don't see the code. Code Review is an important part of ny software development (open or closed).