No, that answer on stackoverflow is a **naive **implementation of the same (Sattoro) optimization if I read it correctly.

He actually initializes a whole array for the range needed + 1. (1001)

Thus it wastes space for one element.

Sattoro uses space exactly as is needed. (1000)

So, yes, he came up with a solution that in effect is like Sattoro's, but his implementation is not optimal (one may say wrong, but that doesn't do him justice).

My implementation is EXACT Sattoro and allocates space for the EXACT number of elements.

Note Sattoro is indeed a modified Fisher-Yates.