[EMAIL PROTECTED] writes:

| This code has moved from MAClisp and VMLisp to several "evolving"
| lisps and eventually to several "iterations" of CLtL1 common lisp
| as it was evolving. Even when the lisps changed from single to
| dual function namespaces (a deep semantic difference) it was not
| unreasonably hard. 
| 
| Certainly the task is much easier than moving from linux to windows.


The actual problem several of us have been running into and are
pointing out are not that we are moving a pre-ANSI Lisp program to an
ANIS Lisp program.  The the problem is that of the *variabilities* of
existing ANSI Lisp implementations, the insufficient portability made
by ANSI Lisp, etc.

FWIW I've  helped converted large non-trivial programs (e.g.) from K&R C
(actually dialects of K&R C) to ISO C.  I've also been helping 
the existing GCC's ISO C soure base to be *compilable* with ISO C++.
The task is non-trivial, but OK.  Because there are *usefully usable*
C and C++ standards.

[...]

| "A poor workman blames his tools"

How do you qualify the one who refuses to listen and understand what
people are reporting as the problem?

-- Gaby


_______________________________________________
Axiom-developer mailing list
[email protected]
http://lists.nongnu.org/mailman/listinfo/axiom-developer

Reply via email to