Just to be more explicit about the complete logical fallacy here, extrapolation from pre-v1.0 makes absolutely no sense. In the last few months before v1.0 was out, all of the dev discussions were like “okay, we have 6 months left to break anything. After v1.0, they are fixed forever. So what are all of the syntax choices you don’t like? What are all of the naming choices you think are inconsistent? What are all of the weird workarounds you think should be removed? Okay, let’s triage, make decisions, and make breaking change fix all of these right now, and set in stone a good v1.0. After that, our choices will be fully documented for the future to know why this is decided, and we can move on without having to ever argue the name of function X”.
Looking at that and going “they broke a lot before v1.0, therefore they will break a lot after v1.0” is a fallacy that completely ignores all of the reasons for why v1.0 was such a breaking release in the first place.