Re: [Chicken-users] Binding identifiers with set!

[felix]

On Mon, 13 Oct 2003 15:55:10 -0700, Dale Jordan <address@hidden> wrote:

> No need to become lawyerly; The final paragraph of R5RS (5.2.1) clearly
> sanctions Chicken's behavior.  I would prefer it were at least an option 
> to
> require "define".  Documenting the behavior is also good.  (C/C++ 
> standards
> require documentation of the innumerable things left to the 
> implementation
> to decide.  Schemes should do the same.)

Absolutely. The behaviour will be documented in the manual.
Currently, I will try to improve the situations by making the compiler
issue a warning, when a non-local assignment refers to an unreferenced
global variable.


cheers,
felix