Il mar 17 nov 2026, 22:31 Peter Maydell <[email protected]> ha scritto:
> > > I see lots of discussion about whether to stop here (and s/four/three/ > > > above), and leaving off the fourth bullet (or splitting it to a > > > separate patch for later) appeals to me personally. But I can also go > > > along with the final bullet if that is what the consensus lands on. > > > > If the consensus is that the legal arguments for the current policy > > aren't relevant any more, I'd like to have some policy update merged > > rather sooner than later. Without this concern, I think I'm mostly in > > the "it's just a tool" camp. > > > > I don't mind the fourth point and don't expect practical problems from > > it, but if some people do have concerns that would take months to > > resolve, I'd prefer to do without it for now and allow at least the > > uncontentious cases without waiting. > > Yes, I would rather we accepted the uncontentious cases and then > had the discussion about the one where it's less clear we have > consensus, than that we hold all the changes up until we get > agreement on all four points. "Push for all four" has got us mostly > "no change" so far. > I don't think this is the case - there has been discussion on all four points and I hoped to have a little more discussion at KVM Forum, so I didn't even think of sending a PR until now. Dropping #4 is totally fine by me. That said, if we are worried about the quantity of incoming patches #4 is actually the least of my worries, and I would like to understand a possible way forward for it. I know personally of at least four uses of AI for large contributions to QEMU that people have experimented with or plan to submit: - the QOM cleanups by Alex (posted, RFC, mostly refactoring) - the Nitro Sandbox accelerator (mature, mostly new code) - Windows VBS enablement (just started) - a textual description of x86 instructions for use with the TCG front end (mature, new build time code + changes to existing code) The first one touches all subsystems, whereas the others are confined to i386. The last one is by me, though done in my personal capacity and not for (or paid by) my employer. Perhaps I can post my TCG work so that we work out in the open what these contributions might look like, with a practical example? If people agree, it could be the Nitro Sandbox work alternatively. I think we definitely should have that community-call discussion > on the AI policy so we can see where we are collectively. I also > don't think we particularly need to tie our policy updates to the > release schedule: it's just a documentation change. > Agreed on both points. Before the next call, possibly already tomorrow, I will post the next version of the policy with the final wordsmithing and #4 removed. I will also post the first (and bigger) half of my TCG work if people think it could be beneficial. Paolo
