4 comments

[ 3.1 ms ] story [ 21.0 ms ] thread
MLISP is really fascinating because it's one of the few LISPs to implement M-Expressions, which are something that McCarthy didn't pursue.
Funny how much complexity you have to introduce by just getting rid of the parentheses.

I later learned that MLISP functioned as a preprosessor for a normal lisp. Don’t know about the macro facilities, perhaps someone can chime in with details?

I think that I remember MLISP. I find Lisp languages to be very easy to read, MLISP is more difficult.

In retirement, my recreational deep learning development is now done in the Hy (hylang for web searches) Lisp language that simply converts Lisp (looks like Clojure) to Python AST. Anyway, because of my background I find Hy to be easier to read and write than Python. So, interesting to see MLISP perform the reverse, a non-sexpr version of Lisp.

So this is where MuLISP/MuSIMP - dual language system came from in 1978. It worked quite well. You could easily switch from LISP to "Algol" at anytime without losing much anything.