As a software engineer currently learning test-driven development, it's fascinating to see all the parallels in SpaceX's approach to aerospace. It's well worth jumping back to around T-40 in the stream to hear and see lots of what they're doing.
Really makes you appreciate that you can re-run a failing test a hundred times before it passes at negligible cost, while applying the same ideology to spaceflight means tens of millions a pop at minimum.
