[R6RS] libraries

[dyb at cs.indiana.edu]

> Excuse me. I'm lost! In your preference, who is allowed to modify a
> variable?

My original preference was that an exported variable be mutable by both
exporting and importing modules and that mutability be controlled by some
other mechanism.  More than half of the remaining editors have spoken in
favor of some sort of mutability restriction, however, and I'm willing
to go along with the idea as long as we get out of it the ability to do
some better program analysis.

My current preference, therefore, is that exported variables, both direct
and indirect, be immutable by exporting as well as importing modules,
i.e., that exported variables be truly immutable.  I hope the reasons
I've given are sufficiently convincing to the majority of editors, but
if not, I agree with you that we should at least disallow assignments
outside of the residual (post expansion) code of the expanded module.

Kent