This is an interesting point I hadn't considered before: converting run-time failures to compile-time failures is already desirable on its own, but in the conte

This is an interesting point I hadn't considered before: converting run-time failures to compile-time failures is already desirable on its own, but in the context of submitting a patch/pull request to an existing piece of software, it also results in shifting more of the maintenance burden from the person evaluating the submission to the person submitting, since presumably they will usually not submit their patch until it compiles cleanly. I wonder if easing the maintenance burden of beleaguered open source maintainers is an even larger benefit in practice than catching errors at compile time.