Doug S (foo.delete@this.bar.bar) on July 13, 2021 12:29 am wrote:

> Any programmer with a four year degree who needs TWO YEARS of retraining just to use a different language
> (unless it is a totally different paradigm, like Lisp...and even then...) isn't worth employing.

Lisp is not that different. In fact, I'd transform code of all ALGOL-like languages I see in my brain to some Lispy IL (without "unlimited extents", or GC, for sure) for analysis purpose for a decade.

The basic rationale to allow such transformation is the fact that all these language share a very closely-looked core of operational semantics which is somewhat an extension of lambda calculus (with first-class states, optional typing and some other derivations). To be specific, ignoring the typing parts, this is formally λvC-calculus by Matthias Felleisen (for language without first-class continuations, the C part can be further ignored). The current version I'm using is some refactored homebrew version of the calculus quite similar to vau calculus by John N. Shutt. With this basis, any core programming language designs containing variables, functions and lexical scoping looks like toys to me:)