Re: JACAL, scm

[Thomas Bushnell, BSG]

Tom Lord <address@hidden> writes:

> Briefly, supporting both standards: R4RS (with the choice that '() and
> #f are the same) and R5RS is not the problem.  The problem is making
> R5RS the default -- that decision represents a major change in the
> (admittedly crude) design spec stated at the outset of the project.

As I read the standard, R4RS does not permit that () and #f might be
the same, it says in straightforward language that they are distinct,
in exactly the same language as R5RS.  It then notes that in some
implementations they are not distict.  That seems clearly to be a
warning to users, not a statement that it's perfectly fine for an
implementation to deviate in such a way.

Moreover, in no standard of Scheme has it ever been acceptible for nil
to match either () or #f.  What do you want nil to be?