No, I'm serious. With datasets that small, there's very little performance difference between the two methods, so a bubble sort isn't unreasonable at all. Depending on the circumstances, a bubble sort can be the better choice due to other considerations becoming relatively more important.

In what way?

Especially if there’s little or no performance difference, may as well minimize code size.

Is everyone here feeling a bit... devilish?

I would pick a standard library sort function at that point.

But more tedious and somewhat more error-prone to write by hand.