Re: Bug & co

[Jerzy Karczmarczuk <karczma@info.unicaen.fr>]

Yannis BRES comments Manuel Giraud problem:
 
> > I think the function `reverse!' in MzScheme 103 is buggy.
...
>      I think that the behaviour of functions with side-effects is undefined on
> pairs that are not allocated at run-time :  when l is defined by (define l '(a
> b c d)) instead of (define l (list 'a 'b 'c 'd)), pairs are built at compile-
> time...

Couldn't you have checked before answering? With l defined as (list 'a 'b 'c 'd)
you MUST GET EXACTLY THE SAME BEHAVIOUR. 
How, for goodness' sake, l can guess what happened to the data it points to?
It continues to point to the *same cell*.

This is - sorry to appear harsh - an *elementary* fact belonging to what we
call "call by value" (here: dereferenced argument).


J. Karczmarczuk