3 ms·
OMG what a fantastically obvious idea! Now that you've mentioned it of course ;-) There may be reasons this isn't as simple as it sounds but it's worth conside
by phkahler 9y ago
OMG what a fantastically obvious idea! Now that you've mentioned it of course ;-)
There may be reasons this isn't as simple as it sounds but it's worth considering.
- Maybestring 9y agoNot that it wouldn't increase confidence, but if you don't know what you needed to be worried about, you can't be sure that you really tested it. Loading and unloading fuel repeatedly doesn't validate that the ground crew will be sober for a 5:00am launch on May 6th. Validation testing is good for testing a known unknown. Whether it's good for testing an unknown unknown, is unknown.
- lucb1e 9y agoIt's a good point that indeed there might be something different at actual-launch-time versus testing time, but I'd imagine that doing the whole procedure a few times (as complete as possible) would reveal any important unknown unknowns if present. (Where I define important as 'reasonably likely to occur': if they are indeed so likely, they'll probably occur with repeated tests.)