Re: Testing dangerous change in tomorrow's snapshot (was: Re: Guile segv)

[Mikael Djurfeldt <mdj@nada.kth.se>]

Maciej Stachowiak <mstachow@mit.edu> writes:

> > I've now renamed `(user)' to `(guile-repl)' because of the assumption
> > that, in a future guile, the sets of bindings exported from the
> > modules (scheme), (guile), and, (guile-repl) will have the following
> > relations: (scheme) will be a subset of (guile) which will be a subset
> > of (guile-repl).  (I assume that (scheme) will be the core Scheme
> > language as defined by the RnRS reports.)
> > 
> 
> It's good to have a (guile-repl) with repl-related things defined in
> it and the contents of the (guile) module re-exported, but I still
> don't think the name of the user module should cconatin `repl'. It
> should just be a module that uses all the modules a user may want and
> starts with no other bindings. This will probably include the
> guile-repl module. A user is free to define another module for his own
> use if the default imports of the user module are inconvenient.

I agree.

I made the above change late last night, and in the morning a woke up
with the feeling that I had made a mistake.  :)

Firstly, it is as important to separate user definitions from repl
bindings as it is to separate repl bindings from core guile.
Secondly, it may be nicer to let the repl module be a separate module
rather than a superset of `(guile)'.

So, now I'm going to rename it to `(guile-user)'.

/mdj