absolutely not normal, this is why in my opinion it took them so long to understand the core issue. instead of a nice error message and a backtrace saying something like "failed to parse config feature names" they thought they were under attack because the service was just crashing instead.
I haven’t worked in Rust codebases, but I have never worked in a Go codebase where a `panic` in such a location would make it through code review.
Is this normal in Rust?