I guess we can never expect the tools to find all the mistakes only the straight-forward ones (which seems to be a moving target)...

This could be an example of when things fall through the crack, while rust & its tooling eliminates most of the errors during compile time, it's not entirely impervious...

Overall I'd prefer depending on tooling to catch a subset of errors (which could miss 1 out of 20 times), rather than the added manual review effort with each change...

What happens? We debug and find the error exactly like we would have to if the tooling was “worse”