Note concernée : note/854697

Bonjour,

Les centres de médecine "alternative" ne sont pas catégorisés comme hôpitaux, j'ai donc mis des tags qui permettent aux usagers de ne pas le méprendre pour un hôpital.

Bonjour, la table de picnic a été déplacée loin au nord, est-ce normal ou c'est un loupé ? (C'est commun avec cet éditeur)

Modifications annulées :

changeset/77294081

Bonjour,

Est-ce que ces ajouts ne seraient pas des erreurs ?

Un nouveau message de changeset ! Dommage que ce soit toujours un tag source au lieu d'une description des changements.

Là ça aurait pu être alignement voirie, ajout entrées et passage en multipolygone d'une place.

Maybe you're not familiar with Overpass (Turbo), but the way you want to map is not how it's done in Lyon. If you want I can detail the results of the Overpass query.

If you continue lowering the quality of the map while not answering questions from other mappers yes sadly the DWG will need to be involved.

>You speak about Part-Dieu Servient

Yeah I was talking about Servient - Liberté

>Are you arguing that it should have been kept?

Nope

>1 I do not believe that they in any form break PTv2 or are in contrary to it.

Indeed you seem to have undestood that recently, after this first changeset that sparked this exchange :
changeset/76235549/

>2 I do believe they are more inline with how bus stops are mapped else where in France.

Less inline in Lyon, more inline in France... What about the world, especially recent high-quality changesets ? imo that's what count

>3 Just because a thing has been mapped in one way in the past doesn't mean it has to stay that way for ever, and to be frank, bus stop in general are not mapped right now in any consistent manner in Lyon. And there are older bus stops mapped by others that are mapped just like mine edits.

Exactly, why should we continue using PTv1 or node-only platforms when we have sub-10cm aerial imagery and nice opendata to map them from

>4 I don't see any problem with having two objects, one for the platform (if there is one) and one for the bus stop.

OSM is a database. Having duplicated or useless features is bad design. In turn, I don't see any problem having only one object where the passengers await, and one object where the vehicle stops.

>I've asked before and I'll ask again; is there a wikipage (or some other documentation) about how to the community in Lyon prefers to map? Just to be clear, I don't expect it to be in English.

No, there is not. But as you can see in my overpass link most stops have use PTv2 and have platforms as ways.

>Specifically the fact that you kept deleting data that I've added to the map that you considered "duplicates" and you did that without any kind of communication. I just noticed it by a chance.

I did answer. And there is no need to ask every contributor before modifying the map. Imagine the infrastructure needed to ask each contributor about each modification, tracking their answers, and mapping long later.

Did you ask me (or anyone) before adding this duplicate platform ?

Please answer to my whole messages, just like I did. You missed explaining :

- Why don't you consider yellow zebras, raised curbs, tactile pavings as valid platforms ?

- Why do you want to overwrite the will of the majority of power-users in Lyon, without asking before ?

- Why don't you use meaningful changeset messages ?

- Do you acknowledge that while there are more PTv1 stops (due to being an older way mainly), PTv2 is the dominant way of mapping PT on recent changeset globally ?

Part-Dieu Servient : It seems that those stops are not in service

Saxe-Lafayette, Edgar Quinet, Foch - Roosevelt Metro, Barrème : You removed the platforms that were well mapped for years, and added platforms as nodes, and kept an highway=platform without any tag.

I can't make sense of what value there is to have two platform objets, in addition to using a node instead of ways for well-defined platforms.

In another changeset you said that such platforms are not visible on the ground. I can quite well see the platform for Edgar Quinet (sorry for the proprietary source):

https://www.google.fr/maps/@45.7655733,4.8445923,3a,51.2y,32.24h,70.18t/data=!3m6!1e1!3m4!1sw16fcj2FQf8dHQHwBlDSzg!2e0!7i13312!8i6656?hl=fr

There is a raised curb, yellow zebras, the cycle lane is closed, and all this align with the length of a bus.

Please stop deleting PTv2 features (stop_areas, platforms). You say you are open to dialog but you go on ignoring many other mapper I listed on an earlier changesets.

>First you delete stuff that I've added without a comment

All my changeset have comments, unlike yours. Again, please use the dedicated field for source, and use comments to described what/why you made changes.

>Then you lie that I've deleted something you've mapped

Yes that was a shortcut. You did add a duplicate object ignoring the existing data. And later you did delete the older, more detailed version as a way to make way for your way of mapping.

>Rest of your message

I will repeat myself, platforms are verifiable, they usually have :
- Yellow zebra paint on the ground
- A raised curb, often of different color

For tram stops, in addition they have tactile pods.

Please contact the many contributors (~40 in Lyon) that prefer ways to map platforms, including a lot more "prominent" mappers than me before removing additional data :

https://overpass-turbo.eu/s/NL0

>you should reflect whether you actually contribute positively to OSM or not

Please refrain from personal attacks.

Hello,

bridge=yes is not necessary on man_made=bridge, only on highways accross the bridge :

osm.wiki/Tag:man made=bridge

Bonjour,

Lorsque vous éditez OpenStreetMap via Maps.me, les modifications sont visibles par tout le monde. Je pense qu'ici c'est un "favori" qui n'aurait pas dû être partagé.

J'ai annulé les modifications ici :

changeset/76612004

> hw=bus_stop is the older one

Being older does not equate being better

> and the one that is more used.
As it is older... And both are getting closer every day : 1.5M vs 2.5M

>The word 'platform' is a poor choice to signify something that is actually a 'bus stop'.

The main goal is to have a unified stop architecture between transport modes. Platform is adapted in those clearly demarcated cases.

>From now on I will map bus stops in the same way as in Paris. In the future if you find something lacking in my mapping I would appreciate it if you contact me first before removing any of my mappings

No I will not contact you first. Did you contact me first before duplicating then removing what I mapped in better detail 5+ years ago ?

I will revert to the existing, more precise situation in Lyon. Feel free to map in Paris if it better suits your views :)

On a side note : when you upload changesets, please use the "Source" field in JOSM to indicate the source, and fill out an explicit message for other mappers.

> For the first the corresponding bus stop across the street is mapped

I didn't quite get this, do you mean this one ?

node/6907031066

If so, this is exactly the same situation.

I understand your concern, as some bus stops platforms can be minimal : a pole and nothing else.

In this case, there are road markings and a raised curb, so there is a platform : https://www.mapillary.com/map/im/DNEK6YWNEsqCyY6WrWMXCw

Maybe you don't want to go the extra effort of mapping platforms in those cases and that's fine, but please don't duplicate the entities.

>It doesn't render in most maps

This should not factor in what method of mapping. For a while casinos where not diplayed in "most maps", so should have they been tagged as tourism=attraction to be displayed for the time ?

> I'm using Osmand to find by way around Lyon, and if hw=bus_stop isn't added one can't really find the closest bus stop on the map

I suggest reading the following article, straight from the horse's mouth : https://osmand.net/blog/guideline-pt

I don't feel this way, but you're not alone regarding addr:street vs associatedStreet. I think the best compromise for you is to use the existing assicatedStreet if available or use addr:street if there is none.

I think it's better not to mix the two mapping scheme on a street.

Hello,

There is no need adding addr:city, addr:country, and on nodes part of an associatedStreet relation addr:street. All this data is already mapped as boundaries and used by most programs, such as Nominatim (the osm.org search engine), OSMAnd and Maps.me (most used mobile applications).

An example :
node/1883877946

Hi, sorry for the french changeset message, which translates to :

"Remove duplicated bus stop"

As you can see in this changeset's modified nodes, I've deleted the node you recently added to merge it with the existing one I mapped 6 years ago.

There is no special rule in Lyon, and nothing against you. I'm just trying to avoid letting duplicates in OSM.

Bonjour,

Est-ce que les anciens bâtiments sont toujours là ? Actuellement ils sont les uns sur les autres sur la carte.