On: Why bad scientific code beats code following "best practices"

[htns]

Yeah. Most scientists are doing mathematics, not programming. Somehow math has done just fine without developing the patterns you see in programming. It's as if they weren't necessary in the domain.