Hi Gabriel, Gabriel Wicki <[email protected]> writes:
> I disapprove > > > > I have taken my time to reflect and meditate over the proposal but I am > unable to find my peace with it. That's very brave of you for speaking your heart, in the context of this GCD which has been passionately debated, to put it mildly. I abstained from letting my personal decision out not only for being jet lagged and missing the deadline but also from the pressure of wanting to reach consensus (I appreciate the hard work everyone has poured in it). If I had done so, it would have been at best a "I reluctantly accept", for the reasons I and others have expressed during the discussion period [0]. [0] https://lists.gnu.org/archive/html/guix-devel/2026-06/msg00244.html Per our GCD, disapproving means that more discussion and thought is needed before ideas in the GCD are accepted by the community. Which is not the worst result; we have already a much better understanding of where we stand to iterate on the next GCD, or what I think would be simpler and work better, propose very short and concise new LLM-related guidelines to include into our user manual via a patch (PR) and bringing people's attention to it via guix-devel that can hopefully be agreed without much back an forth to address what I think is the most serious issue posed by the use of LLM authored changes to the Guix project, the legal risks caused by LLM's output copyright uncertainty. As a stop-gap measure, we can continue (well, we haven't really had to block anything obviously LLM-generated, vibe-coded yet, but I think we agree that we wouldn't let this kind of change in) blocking LLM-authored changes until such explicit guideline is written down for all to refer to. Happy hacking, -- Maxim
