Bonjour VELODB,

Merci pour tes modifications. Je ne comprends toutefois pas pourquoi sur un segment de l'Avenue Jean Moulin
way/1209963986 le tag highway a été effacé. Serait-ce une petite faute de manipulation ?

Bonne continuation,

Patchi.

Le seul giratoire de ce changeset est à Carpentras osm.org/#map=19/44.060332/5.044558.

Bonjour Jean-Louis;

Il va falloir que tu m'aides un peu sur ton commentaire. Tu parles de quel giratoire car la zone est assez grande.

Patchi.

We split only when there is a relation (which is per definition only using a portion of the roundabout). Therefore when you don't have a relation (which is the case for the Chemin du Bois de Cornage, you don't need to split the roundabout segment more than it is now.

The split of the 5 roundabouts is now done.

Hello Meinew,

No trouble about this, this is a community. :-)
I already revert one roundabout and will do as much as I can today. I will also check the bus route relations in the coming days so no worries it will be OK soon.

Best regards,

Patchi.

Hello Meinew,

Splitting the roundabout may be an purpose and was indeed in this case. We discussed a lot about the best practise for roundabouts in the french OSM community (see the feed here https://forum.openstreetmap.fr/t/cartographier-les-rond-points/24548/15). Please check the next time you want to have one single roundabout object if there relations onto this roundabout - and if yes please let the roundabout as it is. This could save us a lot of time to correct the relations.

Thanks in advance and best regards,

Patchi.

Hello HaPe-CZ,

the problem that you raise is related to a renderer. Instead of resolving this problem by changing the renderer you use for your purpose, you are skewing the data by removing a part of the train platform where the transport shelter is (the shelter in this case is still onto the platform). This results among other things to new visibility problems on other renderers. Take the change you made in Saint-André-les-Alpes with the official transport layer of OpenStreetMap changeset/152330529#map=19/43.96985/6.50669&layers=T. Now it seems there are 2 platforms and a building in the middle. What should your general user think in this case?
What I also do not understand, shelters on platforms are displayed with the most standard layer Mapnik with a POI.

Best regards,

Patchi

Hello HaPe-CZ,

Thank you for your answer. As you may read on the OSM wiki osm.wiki/Tagging_for_the_renderer, tagging for a renderer is not a good practise.
So the question is not if the 2 elements are displayed (through an renderer such Mapnik) or not but if it makes sense to put a shelter as inner way from a platform multipolygon. As I already wrote this makes no sense to my humble opinion, as the shelter is onto the platform (as other 'objects' as bin, bench which are also part of the public transport platform). The multipolygon suggests that the shelter is not a part of the platform.
Moreover the platform way still belongs to several train route relations with the role platform, but as the platform way itself (which is an outer way from your multipolygon) must be left untagged (see osm.wiki/Relation:multipolygon) this cause inconsistencies into these train relations. Putting platform tags back onto the platform way is not the correct solution as long as this way is an outer member of a multipolygon (the outer way must be left untagged). To me the only correct solution is to put it back as it was before und delete the multipolygon.

Best regards,

Patchi.

Bonjour chamdam,

Merci pour ce changement. Toutefois l'ajout des tags sur le chemin délimitant la plateforme ne résout malheureusement pas le problème initial (voir mon commentaire sur le changeset changeset/152324477). PTNA ne donne certes plus d'erreur mais le problème a la base est la transformation de la plateforme en multipolygon.
De plus dans une relation de type multipolygon, le chemin du contour externe ne doit pas être étiqueté (voir la page wiki multipolygon -> usage osm.wiki/FR:Relation:multipolygon#Utilisation). Je serais donc d'avis d'attendre un retour de utilisateur HaPe-CZ avant de faire les autres changements nécessaires.

Bonne continuation,

Patchi.

Hello HaPe-CZ,

I don't quite understand why you change the platform ways from the train stations Saint-André-les-Alpes (way/395252564) and Entrevaux (way/396734929) into a multipolygon.
First of all, the platform ways are used into several train route relations and therefore we do have now some inconsistencies. The former platform ways with the role platform are no platforms anymore (just an outer way from a multipolygon).
And secondly, I doubt, that the shelters which are now used as inner ways for the new created multipolygons are not onto these platforms. What is the problem to have a shelter onto a train platform? In my humble opinion it is not necessary to create a multipolygon in that case. There all kinds of objects onto train platforms (bin, bench, etc…) and a shelter is one possibility. In this case the shelter is still a part of the platform. With your changes, it suggests that these shelters are not a part of the platforms anymore.

Best regards,

Patchi.

Hello HaPe-CZ,

I don't quite understand why you change the platform ways from the train stations Saint-André-les-Alpes (way/395252564) and Entrevaux (way/396734929) into a multipolygon.
First of all the platform areas are used into the train relation and therefore wo do have now some inconsistencies. The former platform areas with the role platform are no platforms anymore (just a outer way from a multipolygon).
And secondly I doubt, that the shelters which are used as inner ways for the new created multipolygons are not onto the platform. What is the problem to have a shelter onto a train platform area? In my humble opinion it is not necessary to create a multipolygon, there a several objects onto train platforms and a shelter is one possibility. In this case the shelter is still a part of the platform. With your change to multipolygons it suggests that these shelters are not a part of the platforms.

Best regards,

Patchi.

Bonsoir Jean-Louis,

Je comprends ton étonnement. Toutefois je n'ai fait, dans ce cas, que corriger la cohérence de la ligne 1 sans remettre en cause les changements (les arrêts de bus n'étaient pas tous au bon endroit dans la relation PTv2).
Donc je suis bien incapable de te donner plus de précision quant au prolongement de la ligne 1. Il va falloir voir le changeset changeset/152000912 et poser la question à l'utilisateur @adni978.

Cordialement,

Patchi.

Bonjour wisi_,

J'ai du mal à comprendre l'ajout de pistes cyclables (par exemple way/1273080638) à côté d'une voie verte déjà existante. De plus selon la page Wiki consacré au vélo (osm.wiki/FR:Bicycle#Voies_vertes) une voie verte est devrait "généralement [utiliser] l'attribut highway=path qui correspond le mieux à la définition de la voie verte". Donc pourquoi rajouter des pistes cyclables supplémentaires dans cette zone ? Pour moi clairement des doublons.

Bonne continuation,

Patchi.

Hello Tag Upgrade Bot agin,

I don't think the change of the network TECX to TEC was the correct move. Especially it there are references as ref:TECX=* and route_ref:TECX=* in the objects that you change, it should clearly indicate that TECX was the right network name. Even if I appreciate the other art of changes you made (change old tags, add wiki tag, etc...), I think you shouldn't change or add the network name as several of your last 'network' changes were not entirely correct and sometimes even false (like probably this one).

So please take this comment into account and try to adapt your rules to avoid the network changes if you are not 100% sure of what you do.

And please complete your OSM profile in order to give more details about what you do as a bot. Furthermore as a bot you should be found into the official bots list osm.wiki/Bot.

Thank you for your comprehension and best regards,

Patchi.

Hello Tag Upgrade Bot,

This change doesn't seem correct. Changing the network from the train relations und master relation may not be a good idea. More over I don't understand why add an network=* and network:wikidata=* to the train station? So please update your rules.

Thanks in advance and best regards,

Patchi.

Hello Tag Upgrade Bot,

This change doesn't seem correct. The TER 04 ist not a member of the network TGV InOui and never will be.
Therefore please update your rules in order not to change the
`network:wikidata=Q3512123` and
`network=TER Provence-Alpes-Côte d'Azur` when you have a name like TER.

Thanks in advance and best regards,

Patchi.

Bonjour Monsieur GARAYT,

Quelle est la raison d'avoir transférer pour la relation TER 05 (relation/976032) "network=TER Auvergne-Rhône-Alpes" à "network=TER Provence-Alpes-Côte d'Azur" et "network:wikidata=Q41585492" à "network:wikidata=Q3512123" ?

Car le tag network:wikipedia=fr:TER Auvergne-Rhône-Alpes n'est du coup plus vraiment cohérent avec le reste. De plus il existe une ligne TER 10 (relation/4502335) pour les TER Provence-Alpes-Côte d'Azur. A mon humble avis il s'agit d'une erreur, cette ligne de train devrait être sous le réseau TER Auvergne-Rhône-Alpes.

Bonne continuation,

Patchi

Bonjour Jean-Louis,

Le détail est toujours difficile à obtenir. J'ai découpé plein de surfaces dans cette zone (les premiers imports Corine ont créé des surfaces gigantesques qu'il a bien fallu découper et réduire). Différencier une forêt d'un terrain agricole c'est faisable. Et même si ça peut bouger au fil du temps c'est relativement stable. Différentier les cultures permanentes (vergers, vignobles d'un coté - culture classique de l'autre) comme tu l'indiques est sans doute faisable et stable également.
Après faire la differentiation de cultures par parcelle, je trouve que ce sera difficile voire impossible à mettre à jour. Non seulement il faut avoir des images aériennes précises et récentes. Nous avons déjà beaucoup de chance avec les images IGN mais celles ci ne sont actualisées que tous les 2-3 ans. La zone du Ventoux date de mai 2021. Donc à moins d'avoir une armée de volontaires sur place capable de relever sur le terrain ce qui est cultivé, je trouve ceci bien difficile.
Autre exemple pour la difficulté dans le détail est la forêt. Ce n'est pas toujours super clair. Sur les pentes du Ventoux en hauteur certes il n'y a plus que des résineux mais plus bas difficile d'avoir une limite claire à partir d'où nous avons une forêt mixte voire que de feuillus.
Je pense que ta méthodologie est correct. Commencer par ce qui est clair avec les vues aériennes et avec quelques retours direct sur place faire les petites corrections nécessaires.

Bonne continuation,

Patchi.

Bonjour Jean-Louis,

parfait !
Si jamais tu as besoin que je fasse quelques changements fait moi signe.

Bonne continuation,

Patchi.