Hi Maxim, Maxim Cournoyer <[email protected]> skribis:
>> I think what you describe is a loophole in the process: because whether >> comments were “constructive” is somewhat subjective, it’s hard to reject >> a disapproval on this ground in practice, which threatens the process. >> We should look into fixing that. > > I suppose this is how it can look depending on your perspective, but > from mine, there was no fool play in the disapproval of either GCD 007 > or 008. As we’ve seen, opinions differ. What I’m bringing up here is that the clause saying that disapproval may be accepted only if there were “constructive comments”—which we considered key when the GCD process was drafted—is almost inapplicable in its current form. That’s why I view it as a loophole: there cannot be consensus building if, in practice, one can come up and cancel the whole process. > One thing I'd like to raise is that there was some kind of pressure > conveyed by some proponents of the last GCD that made it feel like a > question of survival for the project, which of course raises the > stakes and puts an unhealthy amount of pressure and focus on the > people that had disagreements with the GCD. I'd like if we could > avoid that in the future. There are two elements here. First, the pressure to not disapprove a GCD, any GCD: yes, that pressure exists, by definition. The GCD process is *not* a voting process; it’s a consensus-building process. Disapproving is throwing two months of work by many people to the trash. It’s a powerful and necessary thing, but with it comes a huge responsibility. It’s a last resort. As for the pressure specifically on GCD 8… You’ve probably noticed that by now every major free software player has adopted a policy towards LLMs, sometimes lenient (Linux, systemd, etc.), sometimes amounting to a near-ban (GCC, Zig, Codeberg, etc.). This should come as no surprise: writing software is the core of what we’re all doing. I know of two “no policy” projects: Guix and Emacs. “No policy” means any committer can accept or reject contributions based on personal preferences. This fuzziness can only lead to friction among contributors. I think it’s irresponsible. (I would have more to say about the pressure insinuations but I’ll keep that for another day.) Ludo’.
