ReL Unix implementations, was: Who wants a nice monitor (UK)?

On 7 Feb 2012 at 22:47, Eric Smith wrote: I suppose it depends upon how you look at it. There are plenty of machines out there with incomplete (i.e. not all instructions present in hardware), where execution of the unimplemented instruction causes a trap to a subroutine that performs the deisred operation. I would not, as a whole, call programs running on those machines, "emulated", even if they use some emulation facilities. Put on the appropriate fully-implemented hardware, nothing is emulated. Whle you could claim that 6502 SWEET16 is an emulator of sorts, really, the difference between a subroutine call with a variable number of arguments and SWEET16 code is vanishingly small. Both are activated by a JSR and control eventually returns to the calling program. So, back to the original premise, operations that take several bytes on an 8080 could be reduced considerably by resorting to a SWEET16 sort of implementation--i.e. an RST byte followed by an opcode byte, followe by operands, if any. So long as control returns to the calling code, that code, as a whole is not being emulated. Contrast this with P-code implementations, or CPU emulatons, where the program never gets control of the real P-counter--it's always being managed by the P-code interpreter or the emulation package. --Chuck