> There is no way that much gas at ~4 bar is safe. It would destroy all nearby buildings if it failed. Wat? This is a launchpad, the safety assumption at play

> There is no way that much gas at ~4 bar is safe. It would destroy all nearby buildings if it failed. Wat? This is a launchpad, the safety assumption at play here is that this vehicle (and in fact a much larger vehicle with this vehicle as a second stage) can be filled with liquid methane and liquid oxygen, explode (with the energy of a small nuclear bomb, a few kilotons), and no one will get hurt. The area around the launchpad is kept clear during testing by local law enforcement for this very reason. Hours before the test that failed, in a different test, the same test article was pressurized with room temperature nitrogen, likely up to ~7 bar (based on previous tests for which we were given figures). For comparison, this is the same location that the following test was performed at, all the additional structures you see have been added recently by SpaceX as ground support equipment and they are undoubtedly perfectly aware and ok with the idea that they might be damaged in the case of a failure https://www.youtube.com/watch?v=bYb3bfA6_sQ https://www.youtube.com/watch?v=bYb3bfA6_sQ As for the rest of your post, yes, if you assume the people running the tests were incompetent then they would be incompetent. We can't fully rule out the idea of course, but it's far from the only option. Most likely the failure was not nearly so simple as "we forgot to pressurize a tank" - particularly considering that the bottom tank was clearly pressurized as it was venting. The most reasonable suggestion for a proximate cause I've seen, for instance, is that the pressure release valve on the lower tank iced over and leaked until the pressure was no longer enough to support the above mass and it failed. This is supported by the evidence that we know they were having valve issues related to cryogenic liquids earlier in the day, and that the bottom tank was inexplicably venting for roughly 13 straight minutes before the failure. It doesn't quite line up with Elon's suggestion that this was possibly a test configuration error.