Eli Barzilay scripsit:

> That distinction forces the above to work, but that's little comfort
> for any piece of code that I write, since that disguised assignment
> can break my code completely (unless I invent my own module system,
> and avoid relying on anything but the standard core...).  A module
> system is a much better solution that keeps things sane without
> resorting to such special cases.

R6RS provided a module system at the expense of not providing semantics
for a REPL.  R7RS, however, is mandated to have both.

In practice, a REPL semantics can't be made perfectly consistent.
For example, no Scheme REPL allows post hoc redefinition of syntax forms,
although SCM allows it if the code containing the form has not yet been
executed.  Post hoc redefinition of free variables does work
everywhere, though.

Similarly, for a REPL to be functional it must be meaningful to re-import
the revised version of a module.

--
John Cowan  cowan@ccil.org  http://www.ccil.org/~cowan
Does anybody want any flotsam? / I've gotsam.
Does anybody want any jetsam? / I can getsam.
        --Ogden Nash, No Doctors Today, Thank You

_______________________________________________
Scheme-reports mailing list
Scheme-reports@scheme-reports.org
http://lists.scheme-reports.org/cgi-bin/mailman/listinfo/scheme-reports