[Maakuth]

I'm not very worried about the safety aspect, as the safety features are probably among the most explicit parts of the design. These are carefully designed, modeled and simulated and are also the focus of the regulators. Moreover, there hardly has been a lot of room for commercial nuclear designers to the software development-like approach of trying, failing and trying again. Maybe the early experimental and weapon related reactor designers worked like that at some point? These days safety is the first priority for designers and operators alike. One example illustrating the thinking is the EPR core catcher: while the active safety features in EPR mean it's possibly the safest reactor type ever built, there are provisions to minimize enviromental release of radioactive materials in the case of catastrophic failure.

[whizzter]

I think such design things overall is what's needed to make things safe for the environment all around, but if there was stuff that was known to designers 40-50 years ago that wasn't properly logged "and lived within the walls" then we might encounter minor failures (that in worst case could lead to major ones).

Reading about many of the early accidents in military and early nuclear industry, I'm not too sure that there wasn't quite a bit of "trying and failing" going on back then, and while that culture is not up to the NASA standards we see in space flights we've also seen how some of their contractors got slower and slower to the point where they lost the ability to finish things in time (and as time drags on the repeating of mistakes will come up again).