¡Hola!

Gracias, me alegro de que hayas encontrado este problema. Ya está arreglado (en JOSM), Petilla de Aragón también forma parte de la Región Estadística del Nordeste.

osm.org/changeset/164921581

Javítottam Lászlóra!

Hello all,

Some other occurrences were left in the database (e.g. landuse, building, track, road node).

Now, they are all removed. e.g. osm.org/changeset/157283889

Geometry was modified to mini-roundabout according to the discussion. Thank you for the inputs!

Agreed, it can be remapped so

Hi kitsee,

Based on the satellite imagery, it is an edge case (considering osm.wiki/Tag:junction%3Droundabout description). The give way positions imply that it can be a roundabout.

But there is no shadow clearly marking the one-way/roundabout sign position inside the roundabout, or at the island at the forks. So the designated roundabout nature was not crystal clear.

If you can confirm (by local knowledge) that it is a proper roundabout, it can be retagged of course.

Bonjour Marc,

Je vois ce que vous voulez dire, et vous avez raison de dire qu'il y a plusieurs possibilités de cartographie dans OSM, où l'« utilité » ou l'« importance » d'un feature n'est pas évidente. Par exemple, en ce qui concerne les limites cartographiables (osm.wiki/Key:boundary), je ne me suis jamais demandé si j'habitais dans tel ou tel district de police, de santé, électoral ou diocèse, tout comme vous l'avez écrit pour les régions statistiques NUTS.

Mais c'est là l'avantage d'OSM. Chaque cartographe peut se spécialiser dans le "feature" qu'il aime cartographier et qu'il considère comme importante parmi les nombreux aspects de la carte holistique. Jusqu'à un certain point, les meilleures pratiques d'OSM sont appliquées.

Quoi qu'il en soit, l'ajout de ces régions statistiques restantes n'a pas signifié une importation du tout. Aucune nouvelle géométrie n'a été ajoutée à la base de données. Seules les "ways" existantes ont été combinées en relations, sur la base de l'orientation visuelle des sources données.

Hello jefle,

It was added as a relation, seen as a practice in some countries, that even relatively small islands (without further elements, e.g. landuse multipolygons inside) were rather treated as island multipolygons.

But after multipolygonizing several such islands, this practice was rather discouraged by other mappers.

I think the guideline shall be the multipolygon and place=island documentation of OSM wiki. It is not 100% black & white in this question. But whatever applies better to this island, it can be applied so.

Bonjour Marc,

Le site de Eurostat est disponible ici: https://ec.europa.eu/statistical-atlas/viewer/?config=RYB-2022.json&mids=BKGCNT,NUTS2,CNTOVL&o=1,1,0.7&ch=TRC,NUTS&center=46.52911,9.25134,6&lcis=NUTS2&nutsId=CH01&

Et Wikipedia est ici pour un comme référence de recoupement: https://en.wikipedia.org/wiki/Lake_Geneva_region

Les régions statistiques de NUTS (niveaux 1 et 2) ont déjà été cartographiées dans 80 à 90 % des pays. Seuls les éléments manquants ont été ajoutés.

Hello woodpeck,

The background is the consolidation of data. Already several one way element islands were already treated as relations (e.g. osm.org/relation/965894 ;osm.org/relation/963182 ; osm.org/relation/543205).

I see what you mean now. In those cases where only the way was present (without existing multipolygon relation for the island), JOSM carried out all the copying.

In those cases you mentioned, the standalone admin_level=* tag left on the way was put into the relation role. As it was the practice in case of other islands (where only natural=coastline and source=* was left on the way). Some examples: osm.org/relation/17825226 ; osm.org/relation/4143917 ; osm.org/relation/3591870 ; osm.org/relation/6914328 ;osm.org/relation/1815263 )

Examples (best practice?) were checked, as I haven't seen exact statement on this (osm.wiki/Tag:place%3Disland ; osm.wiki/Tag:natural%3Dcoastline ; osm.wiki/Relation:multipolygon ; osm.wiki/Relation:boundary)

If there is a consolidated approach on this and they should be kept on the way instead of the multipolygon relation, I'm glad to put them back.

For other islands, where multipolygons were already used, the common practice was having all tags on the relation (e.g. name, place=island, wikidata, etc.), except the natural=coastline, which is kept on the way by default

JOSM's multipolygon creator actually function this way

Hi Aleksander,

admin_level=7 was already on these ways (in many cases without its boundary=administrative pair tag)

During multipolygonization, all these values were kept, as it might had been put there with a purpose of the original editor.

If they doesn't fit according to someone's local knowledge (of the Greek admin system of such small island), they can be challenged of course

Maxspeed=30 kph is the proper comment

It is an interesting question this way, how the "dictionary-like" usage of Wikipedia and Wikidata counts, what sus242424 pointed out. Aka. checking the further name variants of already mapped features of OSM, one-by-one manually, not by any importing mechanism.

Sensum stricto, mcliquid is right, Wikipedia shall not be used in any cases as a source, according to the red box. No further question then.

But sus242424's comment also worth the consideration. As these features are already there in OSM, there is no geodata import in these cases. But rather information cross-check, according to mostly already linked Wikidata and Wikipedia pages, where only a foreign naming standard is checked.

Hello mcliquid,

Thank you for the feedback and the clarifications!

You are right, the bounding box of changes should be definitely smaller, corresponding to lower administrative units or even individual features in this case. Try to focus more on that in the future!

Wikidata as (already linked to each of these nodes) source was unfortunately missed from the source field. But I'll make sure to double-check the sites, you cited.

Thanks again!
osminng

I guess it is clear now. Deleting the original was correct as that line not exists anymore

And the ferry from Gibraltar to Tanger-MED (which is a new route?) should be drawn from scratch.

Hello maiwolf,

On the operating company's website I found.

https://booking.frs.es/#/en/book/xEHjsvOGNHCk8Zw1opf3nMbabHBQswyZ7mihWIXQFG7fW5hoqK47GJpYgkup/outbound-trip

That's why I'm a but unsure because they might simply left it in their database.

But it can be also true that it was resumed since last October

Maybe I try to ask directly from the company

Hello maiwolf,

Could you please confirm (based on field experience) that this ferry line does not exist anymore?

I checked on the website and you can still book Gibraltar-Tanger ferries.

Bonjour Michel,

Désolé pour ce problème. J'ai essayé de vérifier les "relations" lorsque le changement de géométrie a été effectué, mais ce petit problème subsistait alors. Merci de l'avoir corrigé !