Re: [PATCH v4 0/4] refs: run copy and rename through transactions

From: Patrick Steinhardt <ps@pks.im>
Date: Oct 9, 2026, 05:41 UTC
Message-ID: <ash-bs4-WmGuLyCl@pks.im>
In-Reply-To: <CACQ=SRHRp4vKpV5JagAXS1n4iX-LJ7fm78OwMGB5fke2CTC4FQ@mail.gmail.com>

On Thu, Oct 08, 2026 at 09:06:32PM +0200, Maciej Ciemborowicz wrote:

> On Thu, Oct 8, 2026 at 5:46 PM Junio C Hamano <gitster@pobox.com> wrote:

[snip]

> So realistically, my choice comes down to options 1 and 3. And that's
> my moral dilemma: I don't know whether it's better not to submit a
> patch at all, or to do the best I can with the time I have, using AI.

I think one important factor here is the size and complexity of the changes that you are proposing. If the change you're about to submit is smallish and can be reasoned through quite easily then I'd always be in favor of submitting such a patch. But if it's large and complex I'd refrain from doing so because chances are high that there are lots of nuances that the AI will get wrong and that you cannot vet because you lack the context.

And I think that's something that was true even before AI. People that want to contributo to Git have a bit of a ramp-up time, and the guidance has always been to start with microprojects and then work your way up. And I think that guidance continues to apply even when these newcomers are equipped with AI.

Patrick