Object-oriented and procedural programming treat all data as a combination of two things: a place (in memory) and a value that occupies this place. In functional programming we omit the first concept and talk about immutable values directly. This makes programming way simpler because it frees the programmer from thinking in terms of resource and state management and allows them to think in terms of their domain. These advantages carry over from functional programming to functional software architecture. We avoid mutable state no matter the size of the structure.
Blabla und so weiter.
Functional programming languages come with a wide array of
immutable data types. The archetypical immutable data type
is the linked list built of cons pairs.
;; The list 1, 2, 3
(define alist
(cons 1 (cons 2 (cons 3 nil))))
If you want to change the head of alist
from 1 to 99 and keep the rest
as is, you can produce a new list by
calling cons with 99 and the
rest of the old list.
;; The list 99, 2, 3
(define newlist
(cons 99 (rest alist)))