> The resulting breakages will be discovered by automated QA It's a bold move, Cotton. Wiping `Cargo.lock` and updating within semver seems sort-of reasonable

> The resulting breakages will be discovered by automated QA It's a bold move, Cotton. Wiping `Cargo.lock` and updating within semver seems sort-of reasonable if you're talking about security updates like for a TLS lib. "Massaging" `Cargo.toml` and hoping that automated tests catch breakage seems doomed to fail. Of course this was 3 years ago and this person is no fool. I wonder what I'm missing. Did this ever go into production?