It would be an interesting project to split R7RS-small into "stages" and describe them via implementing each stage in the last one. For example: assume R7RS doesn't have `read`, `eval`, `string->number`, or anything else that knows anything about Scheme syntax or evaluation semantics. How do you then implement these in terms of the other operations? This would be a true reference implementation of R7RS, and a useful tool for testing and specifications.
I've had the same idea, but both of you and many regulars on this list are better suited for specifying it than I am.
IIRC the main parts of Gambit were written in R4RS for a long time, to ease bootstrapping?
SRFI 211 (Scheme Macro Libraries) section `The low-level macro facility of the R^4RS` gives rudiments that are enough to build the other macro systems.
