Post

Bozhidar Batsov (a.k.a. Bug)
@batsov.net

One of the many joys of using AI for programming is the creation of huge PRs on complex topics that the authors barely understand, but still suggest "because they work". Here's a great example from #OCaml github.com/ocaml/ocaml/... Kudos to OCaml's maintainers for handling this so gracefully.

github.comDWARF support for macOS and Linux by joelreymont · Pull Request #14369 · ocaml/ocamlDWARF v5 Debugging Support for OCaml Native Compiler This PR adds DWARF v5 debug information to the OCaml native compiler, allowing proper source-level debugging in GDB and LLDB. What's Impleme...
Gary Trakhman
@gtrak.bsky.social

What a colossal waste of time for everyone involved. I use LLMs for code generation, too, but if I get a plausible implementation out of it in a day, I expect to take about 4x as long reviewing it, refactoring it, and making it more robust, and then I still don't entirely trust it.

Bozhidar Batsov (a.k.a. Bug)
@batsov.net

I totally get it. I also use LLMs from time to time, but I'd never commit something or file a PR unless I understand fully the code there. From me the LLM-generated code is just a starting point. Working code is often not the same as good/clear/maintainable/performant/secure/robust code.

0 reposts10 likes

Gary Trakhman
@gtrak.bsky.social

Yeah, I think someone with situational awareness could have presented it as a proof-of-concept, and asked if there's any interest in turning it into a real contribution and what that might take. A PR presented as a real contribution puts maintainers on the defensive and weaponizes their politeness.