[no subject]

[GLS at MIT-AI (Guy L. Steele, Jr.)]

CC: GJS at MIT-AI, CARL at MIT-AI, GLS at MIT-AI, rpg at SU-AI
CC: nil at MIT-MC, rz at MIT-MC

See AI:QUUX;LMFOO > for a rather long proposal to integrate
lexical scoping into the LISP Machine language, and unify
SCHEME lexical closures with the entity/instance/class stuff.
AI:QUUX;LMFOO RMS is a set of comments from RMS on an older
version of the proposal (AI:QUUX;LMFOO OLD), plus my replies.
Many of his suggestions were incorporated into the latest
version.
If some version of this proposal is adopted, it would
effectively make all the things that SCHEME users want to do,
plus the things the actors people want to do, a subset of
the total LISP Machine language.  Thsi would be a great boon.
Also included in the proposal is an analysis of situations
where lexical closures are better than dynamic closures, and
vice versa.