The linked_array crate provides access to arbitrarilly sized arrays
That use the same in-memory representation as traditional slices.
In order to remain performant, this crate depends on optimisations
capable of fully executing tail-recursive loops at compile time.