Re: SRFI package for guile, rscheme, and maybe stalin (minor questions).

[Maciej Stachowiak <mstachow at alum dot mit dot edu>]

Mikael Djurfeldt wrote:
> 
> Maciej Stachowiak <mstachow@alum.mit.edu> writes:
> 
> > Ideally, I would like to see Guile support for as many of the SRFIs
> > as possible, preferrably in the Guile core.
> 
> I trust that you mean adding it to the core _distribution_ and that
> the plan is to _remove_ much of the current code from the core
> libraries (libguile/boot-9.scm) into separate modules in order to
> obtain a less bloated core.

Of course. Sorry if my meaning was unclear. I am not advocating
putting that stuff into libguile or boot-9, and I agree with you about
moving things out. There may be specific examples where we may want to
add stuff though - for instance, SRFI-1 (the list library) defines
some procedures that have the same name as some in R5RS, but an extended
interface (one example would be delete, which has an additional optional
argument to specify the equality predicate to use). In such cases, it
may be better to implement the extended version directly in the Guile
core rather than using the R5RS version but overriding it in an add-on.
But such cases are few.

> Note that Marius and a friend of him are working on a new framework
> for dynamic loading which will be cleaner, more portable, and more
> efficient than the current one.  With this framework having binary
> and/or Scheme level modules which are loaded in on demand will work
> smoothly.

I know. This will be in as soon as I make sure the legal issues are
taken care of and Marius takes care of a few remaining technical issues.

> On my wishlist is a pure R5RS core with as little extra as possible,
> but including important infrastructure, such as module system, object
> system and thread support.

I agree. The dynamic linking changes you mentioned as well as the
first-class environment patch (which I am now reading over) are
essential to making this work well.

 - Maciej