>>105839749
In this case it's applying the rules a + a = a * 2 and a * b + a * c = a * (b + c) until the expression stops changing
read("x + x + x * 2 + x + x * 3 + x")
(+ (+ (+ (+ (+ x x) (* x 2)) x) (* x 3)) x)

fold(read("x + x + x * 2 + x + x * 3 + x"))
(+ (+ (+ (+ (* x 2) (* x 2)) x) (* x 3)) x)
(+ (+ (+ (* x 4) x) (* x 3)) x)
(+ (+ (* x 5) (* x 3)) x)
(+ (* x 8) x)
(* x 9)
(* x 9) can be rewritten (+ (* x 8) x) where x * 8 is the same as x << 3, so later an instruction scheduler might select add r0, r0, lsl #3 if add-lsl is faster than mult

Okay this is an interesting topic. Isn't the old way that every application creates a ~/.application_name directory? Inside are configs, and other data related to this one application. Now the modern way is for every application to put configs in ~/.config and other application data into ~/.local/share.

Can someone explain this to me? Wouldn't it be smarter to keep files from one application in one directory, instead of all configs in one directory and all other application data in another? Because I sometimes want to get rid of all the data one specific application created, e.g. when uninstalling an application. In the modern way, I have to search multiple places to get rid of this. What's the advantage of this design? The only advantage I see is, that it's easier to delete all configs of every program, but keep the application data, but who the fuck would every do this?