Only in old fields with a long history of failures to learn from are engineers expected to be failure-free. You're absolutely right that civil or structural eng

Only in old fields with a long history of failures to learn from are engineers expected to be failure-free. You're absolutely right that civil or structural engineers should not be expected to build a few bridges or houses and hope that eventually they come up with a design that doesn't fall down anymore. Starship is far outside the realm of rote engineering codes that just need to be faithfully executed. They've simulated and tested and validated as much as was reasonable, now they're working with and learning from full-scale prototypes. If it just worked, they wouldn't know how much margin they had. Now that it's exploded, they know a large number of components that were good enough for the phases of flight that it went through, and also the exact stress at which a particular component/system failed: that's incredibly valuable and useful information, and not something you can find on Stack Overflow or an ISO standard.