When was the last time you wrote perfect software? Personally never in my case. The possibility of complex, provably correct software is pretty unlikely I think due to fundamental logical reasons.

I think any ancestor researcher would have at least some bugs in their code.