John McCarthy on Elephant 2000, Lisp, Ruby and the Computer Industry (2008)

infoq.com

26 points·zooso··6 comments

John McCarthy's videos are rare and it's great to see him talking. It's been said that LISP (alternatively J. McCarthy) borrowed the concept of functional closures from APL in mid 60's, that is McCarthy was not the original thinker of closures.

Do you mean PAL, not APL? (PAL's initial implementation was in Lisp, FWIW.) http://en.wikipedia.org/wiki/Closure_%28computer_programming...

My (probably bad) knowledge of history is that the Lisps didn't do lexical scoping particularly well until Scheme. (Hopefully I don't have a false sense of "closures" vs. "lexical scoping".)

Yes I meant APL, not PAL. By functional closures, I tried to mean functions as first-class objects, i.e. ability to pass them as arguments or store them in data structures, etc.

I had read this in an article about Ruby, AFAIK. It was talking about Ruby (or alternatively "Matz") borrowing many principles from LISP and there the author had mentioned Lisp's functions being first-class objects were adopted from APL. I should find that article to clear the things up.

I don't think old APL's had first class functions.

Steele interviews McCarthy, a really nice interview http://www.infoq.com/interviews/Steele-Interviews-John-McCar...

He has such a great command of the subject and of related disciplines like logic and philosophy. It's interesting that he really kind of kept to himself over the decades rather than speaking more.

Reply on news.ycombinator.com