let bindings

[Shriram Krishnamurthi <sk@cs.brown.edu>]

Immanuel Litzroth wrote:

> All this seems rather inelegant to me and not bulletproof (car may be 
> rebound before the define is evaluated), should I be writing my scheme 
> functions like this from now on?

Your observations are dead on.  Don't program like this.  Use the
unit/module systems in PLT Scheme.  Unless you lexically rebind the
name of a primitive (either re-define it inside a unit, or import it), 
both of which are explicit, statically checkable actions, a unit
guarantees that the primitive you name gets the value of the real
Scheme primitive.  You do not need to worry about SET!'s outside the
unit, for instance.

Shriram