🚀 go-pugleaf

In article <1dujcuINNru4@early-bird.think.com> barmar@think.com (Barry Margolin) writes:
> ...
>C's function pointers are almost exactly like the procedure variables that
>many other languages have.  As in many other aspects, C has overloaded an

Indeed, I have always found it helpful to think of procedure variables
in languages like Modula-2 as actually being only pointers to procedures,
since no other operations are applicable to them than those
of pointer variables.

> ...
>case the indirection isn't required.  You can't cast between them and
>non-function pointers.

In K&R C you could (char*), and in C++ you can (void*).

 ...
>they specified the declaration syntax accordingly.  But when you look at
>C++ and its member function pointers, you see that these are more like the
>traditional procedure variables (the environment in this case is the "self"
>object); however, they've kept C's function "pointer" syntax for declaring
>them.

Actually, they are farther away from traditional procedure variables
than are ordinary function pointers.  A "member function pointer" is more
like an offset:  it does not by itself designate an object, and not even a
specific function in the case of virtual functions (dynamic binding
occurs, depending on the actual class of the object with which
the "member function pointer" is applied).

----------------------------------------------------------------------
Markku Sakkinen (sakkinen@jytko.jyu.fi)
       SAKKINEN@FINJYU.bitnet (alternative network address)
Department of Computer Science and Information Systems
University of Jyvaskyla (a's with umlauts)
PL 35
SF-40351 Jyvaskyla (umlauts again)
Finland
----------------------------------------------------------------------