This article is unconvincing--the author doesn't even attempt to justify the claims made throughout. I.e.: > Notice how we treat the endpoints as special cases

This article is unconvincing--the author doesn't even attempt to justify the claims made throughout. I.e.: > Notice how we treat the endpoints as special cases, complicating the simple rule “sum both neighbors of every element.” A better approach is to first transform the input so the special cases vanish, leaving only the general case. Why is that a better approach? The author never deigns to share their reasoning.