> No bug has ever been found in the “released for flight” versions of that code.

I thought that at least his crash was a result of bad constants in flight software: https://www.youtube.com/watch?v=SWZLmVqNaQc

The first comment appears to agree with me.