From: Junio C Hamano Date: Thu, 08 Oct 2026 15:45:57 GMT Subject: Re: [PATCH v4 0/4] refs: run copy and rename through transactions Message-ID: In-Reply-To: Maciej Ciemborowicz writes: > better at it than I am. If I had written all of this by hand, first, > it would have taken me a month, and second, I would have made far more > mistakes than AI does. And thanks to your learning, the next patch series you will write would be of much higher quality. > I understand that the patch is large, and I'm concerned that this > could be a barrier to code review and discourage people from reading Large is not a problem; unnecessarily large is. Code generated by an LLM tends to be unnecessarily large, replete with poor refactorings and outright repetitions of existing code. This often shows that insufficient thought went into crafting the final result. Perhaps this is unavoidable, since they do not really think.