Predicting what a compiler outputs is quiet hard, especially if you enable optimization flags or get near any undefined behavior. But I think you’re right to say that a compiler is more predictable than any LLM. I don’t get the point why you should be unable to audit the output of an LLM? If its used as a tool like described the user should have the knowlage to audit any output

The issue seems to be mostly about the support overhead from really outdated versions still shipped to the public. This seems to be a legitimate concern and why I avoid the distros in question myself.

The issue about “inferior” package managers is not that wrong as well. Version mismatches between dependencies is a big issue in development and why stuff like docker, nix and even atomic desktops are so successful.

I’m a big fan of the classical package managers, especially for my base system but the overhead on developers and packagers is definitely not negligible especially if they use pre release versions of dependencies like bottles did.

Lastly I completely get that you don’t like the tone, the mint/home calendar issue was, while I get the concern, a hard read

Not rust itself, it is a great language with lot of upsides. The problem is the rewirte part. A lot of projects underestimate the burden of rewriting your code from the ground up. A lot focus power on a “new and improved codebase” that is never finished while to old codebase isn’t properly supported anymore due to the focus on the rewrite. Rust is just notorious for these kind of rewrites as ist a fast and memory safe language where a lot of developers see advantages for their project (sometimes rightly so).