On Thu, Jan 7, 2021 at 12:32 PM Michael Thomas <mike@xxxxxxxx> wrote:
If anybody can send a pull request of whatever magnitude and it is pulled by the editors, that's a process problem. That's allowing a backdoor to the process of achieving consensus. Yes it happens now regardless of git, but that doesn't make it a good thing.
But that is also not how it works. First of all, I assume authors are acting in good faith: if there is a major change proposed, they're not going to apply it without understanding how the WG feels about it. But even if they're not acting in good faith, a bad change will eventually be discovered by someone doing a diff or a review. While the "rough" in "rough consensus" refers to the inability to get unanimity on anything, it seems like it also permits a bit of leeway in terms of determining consensus for purposes of efficiency and making progress: based on prior discussions in a WG and the deep knowledge authors have about both the material and the views of others in the WG, some changes are obviously going to be non-controversial, so authors often use their judgment to apply them without actually asking on the list. Bad faith actors and misjudgments will still be discovered.
Note that this is a separate issue from fragmentation/participants not knowing about all the work going on that they might have opinions about. I don't see that as a GitHub workflow or minor change issue so much as an issue of there being so much work that no one can keep track of it all.
Kyle
