Sorry, that was a mistake. These should be separate change sets.
Why does JOSM still not warn about this?

Hi, it's me again:

In the changeset changeset/60430470, I just added a segment to the line you created here so that it is connected to the way in the east of it.

Oh, I just saw that you created way/604737876 in your first changeset. If you don't intend to create a longer route for mountain biking, it's just fine to leave it as it is – no additional relation required. But if it is part of a longer route, you should (or better: you may) add the said relation. (But if it is about mountain biking, better use route=mtb instead of route=hiking.)

relation/103234 is a relation which describes the boundary of the municipality of Arsac. That means, it contains lines ("ways" on OSM language) which go along the boundaries.

You changed its attributes so that it no longer represented the boundaries of that municipality, but just a hiking route around that municipality.

As far as I understand your changeset comment, you tried to create a parcours (route) somewhere there.

The right way to do that would be to create a relation, give it the appropriate tags (as you did with the other relation: type=route; route=hiking, plus maybe a name for the route) and to add the ways which are supposed to belong to the route.

Welcome to OSM! For creating new trails, you must create new relations. Taking other relations and modifying them is not good, because these are gone then for their original purpose.
So I restored this one for now.

You accidentallc changed the attributes of a municipality's relation instead of creating a new one for the route you intended. I fixed that for now.

I restored the previous data now, hope that was ok.

So Ottobiano is no longer a municipality, but a horse route? I fail to understand that.

What was the reason for this change? Doesn't La Chaux-de-Fonds exist any longer as an independent community? "m" as a changeset comment isn't very useful for deciding this.

It's ok now, I fixed it, deriving the actual borders from the neighbouring quarters, especially Champagnat (relation/8412628).

Alas, relation/5970232 does not form a closed ring now. It should be closed at the north end, which it isn't.

What on earth ist this supposed to be?

Alas, this change now makes some parts of relation/3792880 overlap themselves.
I don't know what was the point of doing so, but maybe the coastlines and the boundaries should be separated here.

Something weird happened to the boundary relation of Lorenzo Parodi.

If this quarter (24 de septiembre) doesn't exist any longer (or what was the reason for the attempt to delete it?), it would be helpful to remove it completely instead of just partially. The relation still existst, as well as one boundary segment.

…und Änderung an Stromleitung.

…und zwei Masten ganz woanders präzisiert, danach aber vergessen, den CS zu schließen. Bei JOSM fehlt definitiv eine Warnung diesbezüglich.

war das def:opening_hours:PH falsch?

Note that there is already node/3815077900.

What is the point of the node node/5658269839?