Re: Fixing null-environment and scheme-report-environment

["R. Kent Dybvig" <dyb@cs.indiana.edu>]

> In other words, I'd like the meaning of the argument reserved for its
> intended use, but in no way claim that values other than k need be
> supported.

I quite agree.  As may have been clear from my earlier notes, I think
that supporting multiple versions of the report would be tricky indeed,
and there should be no stigma on an implementation that chooses not
to do so.  The version argument should be considered a check on the
compatibility between code and implementation and nothing more.

Kent