In fact the "computed" bounding box of the changeset that you see here is even larger than the reality (it is rounded by "quadtiles")

As you are working in India, India is in the bounding box of the changeset, but there was absolutely no change in India (or anywhere in Asia).
The computed bounding box of the changeset is a bad indicator here. the affected zones are small maritime areas in North-West Atlantic, in the Caribbeans, and in South-West Pacific where there were minor geometric corrections, and North-Eastern Pacific only for a missing relation member (not for the geometry).

In summary the affected zones are only islands of the French Antillas, Saint-Pierre-and-Miquelon where a a sline lake was updated to include an islet and refine the border, two islands in New Caledonia, the small island of Cliperton and a few tags for islands in New Caledonia (fixing also a terrestrial border between two provinces on the main island). All these in France overseas, but none on continental areas you see in the bounding box. No other country was affected.

No it is not huge, a few nodes updated in Saint-Pierre-et-Miquelon, reference the land-area relations for French overseas from the territorial relations (because of a previous contact by someone that did not find these relations), and nothing else !
Do you know that France covers areas on all continents except asia ?
Only the bounding box is large, but changes are minimal.

My comment was accurate: look at the details on the changeset, you'll see it is in fact very small in terms of data

If you want to complement the claims for other countries, you can do that, provided you don't remove them from French borders. However these competing claims are not effective (and OSM for now looks at the current situation and not possible future situations which may never occur in a predictable time).

All these are official statements from France. There are possibly other claims from other countries, but this French de jure situation is also the one de facto.
They are effective subdivisions of France (including by French law, by French constitution, by INSEE, by national geographic institutes, by the French ministry of Oversas).
It is definitely not a "proposal", all these have been effective since more than half a century (even before the Antarctic Treaty, or before the independance of Madagascar or other African/Indian Ocean countries). Some countries still contest the status of Mayotte and Reunion as well, but they are effective both de jure and de facto...

Yes, but it was listed as part of "TAAF". It was forgotten from "Îles Éparses" only. I've fixed it.

Note: this is discussed on the French OSM talk list.

it started locally (in the Kerguelen islands where most of the nodes and ways are located), then there were many affected dependant relations, I checked many of them and found some of them were broken locally and I fixed them, there was also several edit conflicts on them and to solve them I had to decipher them, and reorder them to make sure nothing was missing and they were still connected.
Thie main reson is that the international border of France covers a large area in the world.
This was not a mistake and it took about one week to solve all problems before uploading sometying that should have taken less than a few hours.

In fact thoese changes affect the international borders of France in one relation that covers most of the world.
Changes were made locally (first in the French Southern and Antarctic territory and then there were errors in dependant relations that impacted French Polynesia, and some parts of the French border around the Sarre river. I did not change anything in South America, except one maritime border in French Guiana, which was related to the change of the international relation for France).
I had to do these related changes to avoid breaking other dependant relations

Also, your threat of dwg is unfair. the current boundaries of "islands" are very approximative (these are only "motus") and none of them are the "baseline" which is the adminstrative boundary for *local collectivities* (but not for the prefectoral administration of France which includes the territorial waters)
There was before this change absolutely no boundary defined at all for communes. I'm ADDING them

Note: boundaries for landareas are not dead, especially not in France.
There are still lot of work to do in French Polynesia, but nothing was removed.
In a later update I'll take into account the "baselines" for communes, but for now the islands are the not the administrative "boundaries".
I've just kept the tag "boundary=land_area" for compatibility, but I know it is not necessary as there's also "land_area=administrative" which is more exact, but present too.
Please don't assume things that you don't know about France.

No: the islands that were present are in false "motus" and not really the islands.
They are not even the adminsitrative boundaries (that were not tagged at all before, because the only thing that was present was the international boundary of France and there was nothing about communes, communes associées and atolls in those communes).
All those changes are reflecting the official administrative status as reported by ISPF (the official statistics institute in French Polynesia)
I did not suppress anything, I added many more precisions, you make false assumptions

I did not include this ID, it was already there,added by others. My changeset only fixed broken borders.
Look further in the history of objects.

Peu importe, mon commentaire en haut est juste, la source Bing aussi, ignore la source du décret qui n'a rien à voir, cela n'impacte pas du tout les données saisies, juste le tag du changeset qui a été "autocomplété" par JOSM (une merde de la complétion automatique même sans rien taper, juste en passant dans le champ avec la touche Tab)

Je crois comprendre: par mégarde j'ai appuyé sur la touche flèche vers bas en éditant le commentaire et JOSM l'a remplacé par un ancien commentaire de son historique

c'est un oubli de modif de source uniqment dans le groupe de modifs, ou bien le groupe est resté ouvert. Bizarre car cela fait un bon moment que je m'étais occupé du Bas-Rhin, là c'est bien une modif pour les sous-quartiers de Nantes. et c'est bien ce que JOSM m'affiche.

STOP tony : tu recommences à casser TOUTES les relations frontières !

De plus il y avait réellement de TRES gros problèmes dans tes modifs car tu avais cassé toutes les relations frontières un peu partout dans la zone ! C'est ce que j'ai rétabli (et c'est **prioritaire** sur les références intracommunales ou même FANTOIR qui sont à faire correctement)

Visislbment tu ne comprends pas comment marchent les relations frontières, j'imagine que tu ne connais pas plus les relations "route" (itinéraires) et relations "associatedStreet" (prises en charge par BANO).

NON. Ce sont des chemins partagés par deux communes, les ref:FANTOIR et ref:FR:commune ne sont corrects que pour une commune, pas pour l'autre.

Donc pas de revert à faire. Les tronçons de voies (way de type highway) ne sont pas le bon endroit pour gérer ça.

Je crois avoir fait les choses correctement en mettant des relations pour chaque commune avec les références, mais si j'en ai oublié (des relations) c'est facile à mettre.