2 ms·
It'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
by lucb1e 9y ago
It'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.)