[R6RS] libraries

[William D Clinger]

Kent, responding to Matthew:
> > In any case, I'm still willing to standardize on the phaseless model,
> > even though I think it's flawed.
> 
> Can I treat this as a change in your vote?  If so, it's a magnanimous
> gesture, which I'll gladly accept.

Before I vote, I'd still like to understand the issue.

SRFI 83 doesn't seem to contain a semantics for the
multiphase interpretation.  In particular, I can't tell
which references are supposed to resolve to which phase
version of the named identifier.

I would guess that the variables referenced by phase N
code that is not part of the code for a procedural macro
would refer to the phase N versions of those variables.
I would guess that the variables referenced by phase N
code that is part of a procedural macro would refer to
the phase N+1 versions of those variables.

I would also guess that my guesses are wrong.  Help?

Will