🚀 go-pugleaf

tanmoy@qcd.lanl.gov (Tanmoy Bhattacharya) wrote:

>_Why_ is it acceptable for 0 to stand for a null pointer constant on a
>machine in which a null pointer is FFFF, _why_ is it understandable for 0 to
>be false and non-zero to be true even on a machine which has `jump _if_ zero'
>as its only machine instruction; and, yet, _why_ is it not acceptable to have
>exit(0) stand for successful exit on an OS where return status of 0 implies
>failure?

A machine that has only one machine instruction (jump if zero)? Doesn't
sound too useful to me. :-)

Of course, the reason that 0 is a null pointer constant and non-zero means
true is historical -- that's what the target machine was like when C
was developed. All other architectures are second class and must jump through
whatever hoops are necessary to allow source code to be copied over,
compiled and run successfully. Likewise, so many programs assumed that
exit(0) was a `successful' exit that the standardizers said `this will work'.

-- 
Andy V (a Digital spoke, not a Digital spokesperson)