Good point. You could also say that simulating two-dimensional arrays with one-dimensional arrays is easier with zero-based indexing: a[x + y * w] rather than a

Good point. You could also say that simulating two-dimensional arrays with one-dimensional arrays is easier with zero-based indexing: a[x + y * w] rather than a[x + (y - 1) * w]. Are there similar use cases that look more natural with one-based indexing?