Hacker News new | past | comments | ask | show | jobs | submit login

I've heard this too. One possible root of this claim is Stallman in regards to Emacs Lisp. I think he made a post where he said that dynamic scoping was faster than lexical scoping for elisp (though this was later shown to be mistaken: https://emacs.stackexchange.com/questions/2129/why-is-let-fa...).

Sorry that I don't have any sources to support this. I can't find the post, and it may have been secondhand info anyway.

I thought he meant lexical scope was unfit if not incompatible with an extensible program. Meaning you had to loosen up the binding to allow for user code to hook in. Lexical scope made this "impossible" or maybe possible but required radically different way to program, probably using faux-monadic style lisp.

Stallman famously worked on specialized LISP hardware at MIT. I wonder if these machines had special optimizations for stack access that made lexical scoping faster on that architecture, and Stallman over-generalized.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact