The language versions are often not translations but interpretations with additions.
The portuguese version also states "when not narrow consider using waterway=canal".
The english version makes the distinction between canal and drain/ditch according to superfluous water vs. useful water.
But anyway "consider using" is a recommendation, not a strict law.

Please use the English version when there are differences:
waterway=drain
There: "For artificial waterways created for carrying useful water or for transport consider using waterway=canal. "
I used ditch/drain for some time before, but when i came across waterway=canal & irrigation=yes I looked at the english version and changed to canal.

In contrary your mass edit of levadas in Madeira is not correct.

These lines were mostly tagged with power=pole, not with power=tower. The latter is usual with high-tension lines and I therefore got warnings by OSMI.
Mid-tension lines (<50kV) are used for distribution, high-tension lines (>100kV) for long distance exchange.
The 60kV lines in Algarve are mostly used for collecting wind park energy and are somehow in between.
By having a closer look the tagging of these lines in this region is very inconsistent: It is a mix of line/minor_line and with tower/pole although of same size and pole/tower distance. In high resolution images it looks more like single poles than (lattice) towers.
Anyway, I have no objections against power=line at 60kV, but this should be done consistently.

Geht im Prinzip schon, nur nicht als AL9, da die Kernstadt keinen Ortschaftsrat hat.
Wenn aber AL10 oder AL11, wäre die Frage, ob nicht Neugreuth auch denselben Status hätte.

This changeset created large MPs with many inners, which tend to be very unmanageable. Furthermore they overlapped many existing areas which made the map totally confusing.
Additionally some outlines were rather coarse when judged by orthophotos of 2021. This is usually caused by machine classification of older and lower resolution imagery. Import even of official imagery should be done very cautious regarding accuracy and actuality.
Above all the licence of the data have to be compatible with OSM terms.

If have not reverted the changeset since some boundaries and areas were more precise or did not exist before.

Hello,
Georgenberg was in fact an underground pipe as part of the swabian volcano almost 20 Mio years ago (Miocene). The jurassic Swabian Alps extended more then 20 km further to the northwest at that time. By erosion of the jurassic sediments the hidden diatreme was carved out because volcanic tuff withstands water better than jurassic limestone.
Georgenberg is not a depression or crater - it has a cone form.
The Swabish volcano consited of hundreds of diatremes, some of them remained as hills, most of them still covered by limestone, some of them came in contact with underground water and made huge steam explosions forming craters (Maar) of several kilometers.
So there are maars to the east, but Georgenberg is a mountain (translated George Mountain)

The name is neither incorrect nor removede, it is at the peak node.
There is no need for an additional place node with the same name.

A place=location name makes no sense when it is the name of a peak (Pico da Cruz) mapped near by.

BTW: You have reverted the whole changeset and thus the road down from the Posto Florestal has got to surface=ground again which simply is incorrect. Some additions like signposts and street names were deleted, too.
It seems you are not aware of the possibility of partial reverts.
But I don't want to rerevert that: I have no interest in nonproductive discussions.

I have been there on 2023-03-22 and at least at that day all people had to use the deviation like the years before.
If you have more recent information, feel free to correct.

Nice to meet you again with your inspiring comments :(.
The Passeio Maritimo in the eastern part of Praia Formaosa leading to the tunnel simply did not exist last week. The tunnel may have been open for some time, but the Passeio has vanished at least the last 4 years.
I left the western part from Camara de Lobos and the eastern part to Doca do Cavacas untouched, but there existed a third route connecting the both via the tunnel.
I just removed there the part through the tunnel and inserted instead the deviation per Estrada Monumental because many people are unsure how to get around the closed tunnel.
If that route is disturbing you, you may delete it, I have no ambitions on that.

Sorry, a single GPS track does not show the details of a trail even after some simplifying. There is an inherent jitter of more than 2 m even for the latest Garmin 2-frequency devices. If you record the same trail on different occasions you will see the difference,
I am using at the moment seven loggers in parallel, three of them with 2-freq-chips. I then do an weighted average of them. The single tracks have an rms deviation of typical 1 to 3 m, but single trackpoints may often exceed 5 m from the averaged track. The accuracy shown by the devices is always overoptimistic.
In short: You cannot rely on the turns from a single GPS track.

correction:
GPS track to OSM way

To me that was just the conversion of something like a GPS track to a GPS way.
Such a high density of waypoints pretends an accuray that isn't true.
Looking at the trail together with raw data in JOSM, e.g tracks uploaded by other users, the now simplified way follows the trend of these GPS tracks rather well. I couldn't identify any missing turns.

There are some slight deviations from arial imagery, but I trust more in a bundle of GPS tracks than in imagery which trends to be distorted in mountain areas.

fixed

Streets and paths outside of settlements are very rarely lit, especially mountain paths. Worldwide you may find only a handful of them, they then should get a lit=yes.
lit=no makes only sense in populated areas where streets and footways usually are lit.
lit=no is usually set by newbies who think that all checkmarks in presets have to be set.
Sometimes you may even find a boat=no.

lit=no is superfluous, here but it leave it as demanded since it does no harm.

There is a large bridge construction across a stream valley. I consists of an aqueduct of a broad canal and an accompanying path also to be mapped as bridge.
Further to southwest there is a small pedestrian bridge (layer 1) over the canal which is now again at layer 0.

This bridge was mapped before unfortunately at the outline of the canal aqueduct construction. Therefore it was in conflict with the aqueduct layer 1.

I have fixed that now.

ok

erledigt

Hallo Peter,

geht i.O.

type=boundary steht in der Relation, ist am way überflüssig, steht nicht im Wiki und ist auch international unüblich.

Habe die tags vermutlich damals beim Präzisieren der Grenze nach LGL ganz einfach von den alten Grenzways übernommen.

ergänzt