This is the mail archive of the guile@sourceware.cygnus.com mailing list for the Guile project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

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


Michael Livshin <mlivshin@bigfoot.com> writes:

> Mikael Djurfeldt <mdj@mdj.nada.kth.se> writes:
> 
> > My position is that the idea of mixing forms of the module
> > configuration language with the bindings it's supposed to manage is
> > flawed.  Using Marius' terminology it's even more clear: The name
> > space manager should not contaminate the name space.
> 
> which of the two things below do you mean?
> 
> 1.  it is desirable to have the ability to conceal the name management 
>     functions from code that doesn't need them.
> 
> 2.  name management should be a separate language from Scheme and
>     never shall the two meet (this looks like a citation.  where is it 
>     from?).

I mean

3. name management bindings should not normally live in the name
   space which they manage

This differs from 1 in that code normally doesn't see name space
management bindings.

It differs from 2 in that name management bindings still can be
ordinary Scheme bindings managed by the name manager.

The paper shows how this works in Scheme48.  RScheme has a very
similar module system.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]