What about non calculus programming?

Question

I recently asked about new languages and language features of the last 10 years, but apparently there is none.

However I noticed one issue is that most of our focus is on calculus of various types. But thinking about it, theoretically I can see Neural Networks and Genetic Programming as non calculus based models for programming, that although lack stand alone implementations, are interesting ideas born in recent years (say less than lambda 1930).

Also equally out-of-the-box are query languages, regex and graph/tree based designs, that although mathematical in part, are not calculus, which got me thinking about what other non calculus designs in programming are important now, and what other fields might provide us with new programming paradigms?

**Note: ** this question is not about "new", but about not originating from calculus.

Answer 1

I think that lexical programming is going to come to the forefront more and more as we get closer to AI that interfaces with people.

Answer 2

Neural networks are more than 50 years old, see e.g. the Perceptron.

Genetic programming is more than 40 years old.

Now the original question got edited, so age is no longer relevant ...

BTW, object-oriented programming originated in simulation, more than 40 years ago (Simula 67).

Most other programming languages are either derived bottom-up from what a machine can do (imperative languages starting with FORTRAN in 1954) or from some mathematical notion of computation (e.g. lambda calculus inspired Lisp in 1957 and later other functional languages. Prolog is another example derived from mathematical logic). The relation of programming languages to calculus that you see is mostly accidental, because numerical computations were the major use of computers in the early days and are still important today.

Answer 3

Calculus isn't math, it's just part of math-- as a whole, all programming languages are mathematically-based, and programs can be turned into algebraic structures (or, more theoretically, into lambda calculus or a turing machine construct). What I'm trying to say is that all work on the computer-science aspect of programming (as tagged) is bound to be based in mathematics, though perhaps not calculus, so the question as such isn't fair. The stuff that isn't is stuff like Knuth's Literate Programming, which focuses on the non-mathematical sides of programming-- documentation, etc., and it's not really new.

Edit: The original question has since been updated. Thanks, I suppose.

Answer 4

New programming paradigms could come about for Parallel Programming practices. Currently, there are many ways of handling programs running on multiple processors, some good, some bad. I expect that with more research in the subject, several core methods will become commonplace for new languages to incorporate so they can be used in an "easy" way.

Answer 5

I think any major shifts in these areas will probably come about through shifts in how we think about the problems rather than the languages we use to solve them.

The tools we have now are rather good at solving computational problems as they are currently framed. New views of the problems will certainly drive the creation of new languages, but that is a by-product rather than the impetus.