Hi Andreas, On Thu, Jul 30, 2026 at 04:54:56PM +0200, Andreas Enge wrote: > On Tue, Jul 28, 2026 at 08:56:08AM +0100, Steve George wrote: > > One thing I've realised since we agreed the GCD process is that a lot of > > ours > > seem to be on deeply emotional and divisive subjects. I'm not sure why that > > is. > > I was expecting GCD's on package naming, instead we've landed up with ones > > that > > cover some of the greatest philosophical, emotional, social and political > > discussions of our times! > > that feels rather natural to me. I suspect that nobody cares about > package naming apart from me, and it is a solved problem - well, if any > issue occurs, we can probably just agree on a reasonable technical > solution, add it to the manual and be done with it. (...)
I agree that's what's happening and you're probably right. I think it's also caused by us adding the GCD process to the community quite late in the project's development so people are used to solving technical problems with PRs/manual. When we designed it I looked at the various examples Simon referenced (and a few others) such as Python and Rust. In those communities their RFC process was started early, they are the focus of any policy change and they seem to have a lot more technical policy. Take the Nix one for example - there's a lot of technical items, along with the organisational: https://github.com/NixOS/rfcs/tree/master/rfcs > We need a GCD when the topic is divisive, with several quite differing > opinions that need to somehow be brought together. (...) Yes, I suppose you're right and maybe we're just exploring divisions that existed anyway and previously there was no way to handle them. Nonetheless, I'm finding the emotive conflict quite worrying and it takes a lot of energy and time to keep up on. > And is this not also implicit in the consensus framing? GCD 001 does not > give a very strong definition of "I disapprove" apart from "opposes the > implementation of the proposal", but there are more details in this > document: > https://www.seedsforchange.org.uk/shortconsensus#agreement > from which, I think, our motivation comes from. > > Block: A block always stops a proposal from going ahead. It expresses a > fundamental objection. It isn’t “I don’t really like it,” or “I liked the > other idea better.” Some groups say the block means "I'll need to leave the > group if this goes ahead". The group can either start work on a new proposal, > or look for amendments to overcome the objection. In cases where the block > stems from a fundamental disagreement with the aims of the group it might be > more appropriate for the individual to leave. > > Implicitly, our interpretation seems to have been that disapproval means > "I'll need to leave the group if this goes ahead". > > I find it unlikely that anybody will feel the need to leave Guix because > they do not like our package naming scheme... So it is natural that > questions with lower stakes are handled outside the GCD process. (This > is further helped by the efforts an author needs to spend on a GCD > and its discussion.) (...) It not explicitly stated in GCD0001 but I agree, that was my understanding of the level required to vote "disapprove". It's a fundamental conflict. It's probably clear enough, but I note I wasn't aware of a formal definition of "consensus" before Guix, maybe others won't be either. If we re-open GCD0001 at some point it would be good to add it. Steve / Futurile
