y := 4x(1-x)
in a typical programming language. The denotational meaning of this is
clear: the state of the store is changed so that the location called
y holds the mathematical value 4x(1-x) after execution. The
operational meaning of this will vary. Under floating-point
computation, this will cause some memory locations to be transfered to
registers, some operations on the registers to be performed, and some
registers to be transfered to memory. Under exact computation, what
actually happens is much more complicated. To begin with, x isn't a
memory location anymore; now it is a ``real number generator'', that
is, a piece of code that generates as many digits of a given number as
we are patient to wait for. Also, for example, the multiplication
operation is now implemented as a procedure that builds a third real
number generator from two given generators.