Re: A module system should resolve, not introduce, name conflicts

[Jost Boekemeier <jostobfe at linux dot zrz dot TU-Berlin dot DE>]

Christian Lynbech <chl@tbit.dk> writes:

> Very exciting. This means we could also separate variables and
> functions into separate namespaces, just as in CL or elisp.

A module system based on first class locations would be really cool --
let's call them "variables".  This would support the generalized set!
entirely:

(define & location)
(define a 1)
(define pointer-to-a (& a))


Jost
:)