Ik heb de Nederlandse claim op Oyster Pond weer terug ingevoerd zodat die weer correct wordt weergegeven, maar daarbij jouw 2023-geometrie van de grens volledig intact gelaten. De overige verschillen met de oude grens zijn waarschijnlijk klein en niet echt de moeite waard om nu ongedaan te maken.

Ik heb ook note-tags toegevoegd om duidelijk te maken dat alvast de 2023-geometrie is gebruikt, maar dat het 2023-verdrag nog niet in werking is.

Zodra het verdrag in werking treedt, zijn nu alleen nog maar hele eenvoudige wijzigingen in Oyster Pond nodig.

The treaty signed in 2023 is not in force yet, because the approval process is still ongoing. There does seem to be some progress this year:
osm.wiki/Disputed_territories#North_America

Also note that once the agreement is in force, there will be a baseline closing Oyster Pond such that Oyster Pond will not be part of the territorial sea (but instead internal waters).

Wegbeheer is teruggezet: https://community.openstreetmap.org/t/vergelijking-maximumsnelheden-osm-vs-nwb-gemeente-amsterdam/143571/45

Ik heb wegbeheerder teruggezet: https://community.openstreetmap.org/t/vergelijking-maximumsnelheden-osm-vs-nwb-gemeente-amsterdam/143571/45

Dat was geen fout, het midden van de weg ligt echt vrijwel op de grens. Ik heb de wegligging hersteld, het is nergens voor nodig om de vorm van de weg geweld aan te doen.

ref=A27 klopt ook niet, want administratief gezien is dit geen afrit van de A27 maar een toerit naar de busbaan. Ik heb noname=yes en noref=yes toegevoegd.

The new route avoids the lower speed limits inside Elten. Previously, the speed limits were not fully tagged.

In deze wijzigingenset heb je mysterieuze tags uic_name=50 en uic_ref=50 toegevoegd aan een aantal wegen (geen spoorwegen), bijv. way/85514791

I've re-added ref=A67 to the carriageways within the service areas; motorway service areas in the Netherlands are considered part of the numbered road.

Ik heb allebei aangepast (inclusief enkele kleine correcties zoals ontbrekende C9-geslotenveklaringen en stukjes maxspeed=100 die waren overgebleven):

* Provinciaal deel van trunk naar primary: changeset/178051719
* Gemeentelijk deel van primary naar trunk: changeset/178052286

Het provinciale deel van de N230 was tot kort geleden autoweg buiten de bebouwde kom maar is nu geen autoweg meer. Het gemeentelijke deel (vanaf Ghandiplein naar A27) is nog steeds autoweg binnen de bebouwde kom. Volgens osm.wiki/NL:Tagging_van_Nederlandse_wegen zou het nu dus half-primary-half-trunk moeten zijn, maar dan precies andersom.

Deze snelheidsverlaging moet ongeveer twee jaar gaan duren. Dat is voldoende langdurig om de maxspeed in OpenStreetMap aan te passen. Prima wijziging dus.

https://www.a27houtenhooipolder.nl/communities/common/images/04fb8cd8-c73f-40eb-a0a4-b00118eb0ca7/Overzichtkaart_gewijzigde_verkeerssituaties.png

Ah, I now see Valhalla treats every admin_level=3 as admin_level=2 by overwriting the admin_level. That explains the observed behaviour.

Okay, I've updated the tagging. The relation now uses the key prefixes reserved:, was:, and is_in: where appropriate.

Regardless, it may be worth debugging what happens in Valhalla because the code really looks like the override you describe should be technically impossible unless admin_level=2.

W.r.t. the former validity of FX: from https://en.wikipedia.org/wiki/ISO_3166-2:FR

> Metropolitan France (the part of France located in Europe) was previously officially assigned its own set of country codes in ISO 3166-1, with alpha-2 code FX, before it was deleted from ISO 3166-1. The code is now exceptionally reserved on the request of France.

This is how it ended up in ISO 3166-3 (mere exceptional reservations in ISO 3166-2 do not lead to codes in ISO 3166-3). The FX code was officially assigned from 1993 to 1997 according to https://en.wikipedia.org/wiki/ISO_3166-3

I'm not sure what you mean by saying FX was never valid. Before becoming an exceptional reservation, it was officially assigned, which is the highest possible degree of validity. As an exceptional reservation, it's only partially invalid.

The Valhalla code you mentioned only reads the value of ISO3166-1:alpha2=* in case of admin_level=2 (on line 194). This relation is admin_level=3, so it will never set the country code to FX, even if you remove the lines of code you referenced that manually set the code to FR.

I believe there are two sorts of practical applications for these country code tags:

1) Find the territory corresponding to a country code.
2) Find the country code(s) corresponding to a location.

Deleting all not-officially-assigned codes may benefit the second category of applications but harms the first. Besides reserved codes this would also affect user-assigned codes like XK.

Would you be okay with adding a reserved: prefix to the ISO3166-1 keys here? That should solve any problems of applications mistaking reserved codes for officially assigned codes, while still making it possible to opt-in to using the reserved codes.

See changeset/176984051

This relation for Metropolitan France should only contain ISO codes that identify Metropolitan France specifically. Data consumers trying to find France in OpenStreetMap by querying country code FR should not find this relation for Metropolitan France (or any other subdivision of France), but only the following relation for the entirety of France: relation/2202162 - that relation already has the FR/FRA/250 codes so these codes are in the correct place in OpenStreetMap already.

FX and FR are currently both valid codes here (with FX having a more specific meaning than FR). While the territory of FX has indeed been merged into the territory of FR, the FX code has not been fully deleted, but replaced by an exceptional reservation. This is why the iso.org website says "exceptionally reserved" as status for FX. If it was truly properly deleted, it would instead say "formerly used" there, like BQ for British Antarctica: https://www.iso.org/obp/ui/#iso:code:3166:BQAQ

Likewise, if you are in e.g. Martinique, both MQ and FR are valid codes (with MQ having a more specific meaning than FR).

There are also other codes like EU for European Union that have always been just an exceptional reservation: https://www.iso.org/obp/ui/#iso:code:3166:EU

In other words, "officially assigned FX" has been deleted and "exceptionally reserved FX" has been added at the same time.

All ISO 3166-3 codes are alpha-4 codes (i.e. codes consisting of four letters) and all alpha-4 codes are ISO 3166-3 codes.

ISO 3166-3 codes are normally created when ISO 3166-1 codes are retired (usually because the territory disappears in the real world). The FXFR code was created because the FX code was 'downgraded' from "officially assigned" to "exceptionally reserved", to offer an alternative for people who were already using FX but don't want to rely on the exceptional reservation (which could theoretically get deleted at some point in the future).

TL;DR: I've reverted this changeset because it incorrectly equates France and Metropolitan France and specifies a four-letter ISO 3166-3 code which is impossible. Also there is a misunderstanding about the nature of the "deletion" of the old codes for Metropolitan France.

Earlier comments: changeset/172975242

Aanvullende bron:

Omroep West (4 maart 2024). Dit is waarom trajectcontrole op A12 bij Den Haag verdwijnt. https://www.omroepwest.nl/nieuws/4812509/dit-is-waarom-trajectcontrole-op-a12-bij-den-haag-verdwijnt

Aanvullende bron:

Omroep Brabant (26 februari 2024). Trajectcontroles op A58 bij Roosendaal plotseling beëindigd: dit is waarom.
https://www.omroepbrabant.nl/nieuws/4426436/trajectcontroles-op-a58-bij-roosendaal-plotseling-beeindigd-dit-is-waarom

Het onderscheid tussen ref A13 en official_ref A4 is hier gemaakt, omdat dit stuk weg formeel-administratief gezien een verbindingsboog van de A4 is, terwijl het functioneel gewoon de hoofdrijbaan van de A13 is (waar het vanaf de onderdoorgang Prins van Oranje ook formeel in overgaat).

De tag official_ref in combinatie met carriageway_ref klopt met wat er op de hectometerbordjes staat (A4 x). Dat je de tags zo moet lezen, staat ook uitgelegd op de wiki: carriageway_ref=*#Netherlands