On Sun, Sep 13, 2009 at 10:26 PM, Aaron W. Hsu <[email protected]> wrote: > On Sun, 13 Sep 2009 21:37:48 -0400, John Cowan <[email protected]> wrote: > A) I tell the implementation where to find it. Easy. > B) I tell the implementation where to find all my code. Also Easy. > C) I put the files where the implementation expects to find my code. > Perhaps not so easy. > > None of these have anything to do with a language. They all have to do > with the behavior of implementations and how that implementation interacts > with the host environment to read in new code. This goes back to something > that I don't think a language standard should define. Maybe a recommended > behaviour standard, or a SRFI, or something, but not in the language > standard that defines Scheme.
Scheme as a "language" has conceded from its earliest incarnation that there are things going on in the implementation that are not explicit in the surface of the "language". It even embraced placing such hidden state under the control of the user through call/cc. This is why I see the view of Scheme as simply a static language as weird. I would put the hidden state of "what library names actually reference" in the same general class as continuation, and an abstract interface to this resource should be provided on the same grounds. Lynn _______________________________________________ r6rs-discuss mailing list [email protected] http://lists.r6rs.org/cgi-bin/mailman/listinfo/r6rs-discuss
