LISP Metalanguage

Recently, someone proclaimed the virtue of S-expressions as "programming
in the metalanguage" (or something like that). This is not strictly true,
although McCarthy's intent was that one would do exactly that.


For those of us who learned our LISP from the 1.55 manual, there was a
metalanguage used there for functions definition, which was known as
"M-expressions". They unfortunately required the presence of the greek
lambda character, which was not available on any keyboard then. When I had
LISP in 1969 (from Bert Raphael) I had to do my notes and written
assignments in M-expressions, my computer work in S-expressions, and (as a
leavening agent) some homework in "dot expressions" (also in the LISP 1.55
manual).


I was told by some of those in the know at that time that the
S-expressions were temporary until LISP 2.0 came out, and then we would be
programming in M-expressions from then on. The intent was to program in
the metalanguage, but it didn't happen. Perhaps Paul Stygar, or John
Allen, or some of those involved with LISP 2 could provide some clarity on
this.


john slimick
slimick@upb.cis.pitt.edu
[It is my impression that once people got used to S-expressions they
realized that the advantages of directly using them was great enough that
they pretty much bagged any metalanguage beyond macros and backquote.
Lisp 2 was one of the early great pieces of vaporware. -John]
--

Return to the comp.compilers page.
Search the comp.compilers archives again.