Openstreetmap-carto: Label public transport ticket vending machines with network

[DerDings]

Expected behavior

Show amenity=vending_machine vending=public_transport_tickets on the map with a useful Label if there is enough space. Labels should be short, so they don't take up too much space on the map
Useful may be:
_What can I use the tickets for?_
_Which vending Machine sells the Tickets for the transport mode I'm going to use?_
This Information is provided by the network= key

Actual behavior

Ticket vending machines are labeled with the text defined as operator=. This gives users the answer on the question _What company gets my money if I buy a Ticket at this place?_ which I don't consider very useful. It also is a very long label in some cases.

Fallback

Sometimes an transport company is not part of any network, so if not network= tag is given on an object, operator= can still be shown instead.

Links and screenshots illustrating the problem

This is an example of vending machines operated by a certain company which sells Tickets for a specific public transport network
Bochum (1, 2)
Freiburg

Additional thougths

May also apply to other vending machines (water, telephone vouchers) that network= is the most suitable tag to label vending machines if it exists.

[jeisenbe]

The combination amenity=vending_machine + vending=public_transport_tickets + network has been used 2129 times: http://overpass-turbo.eu/s/SQe

Compares to amenity=vending_machine + vending=public_transport_tickets + operatorwith 13,486 uses: http://overpass-turbo.eu/s/SQf

The nodes above have "operator=Transdev GmbH" and "operator=Freiburger Verkehrs AG" with "network=VRR" and "brand=VRR", or "network=RVF".

In most cases the "operator" tag seems to include the name of the public transport service operator when used with vending machines, from my prior searches in Wales and a couple others places, but "network" is more specific when it is used differently.

I don't think we could show both "operator" and "network" since this could confuse mappers. Instead, we should drop "operator=" and only show "network=" if this is desired.

[DerDings]

@jeisenbe ~that's exactly what I mean~ EDIT: sorry I got you wrong in first read

[chiak597]

In most cases the "operator" tag seems to include the name of the public transport service operator when used with vending machines

Well, in many cases: public transport operator = vending machine owner (so there might be no sense to tag network and operator at the same time)
However, there are some exceptions, for example parts of Germany where vending machine owner is some local transport company. But that vending machine sells tickets for the whole big regional network (multiple operators governed by regional transport authority).

[DerDings]

I don't think we could show both "operator" and "network" since this could confuse mappers. Instead, we should drop "operator=" and only show "network=" if this is desired.

I don't agree with that.
I think the map should show _features_, not just colored and georeferenced _tags_.

Most users of the map don't know anything about tags and keys and all that OSM stuff. So they will be happy finding the vending machine labeled usefully for them, no matter in which key the text shown on the map is stored.
For _mappers_ who know what tags are and how they are defined, they should map things as they find them. And they should always use tags in the way tags are defined. Mappers should not tag something because they know how it will be rendered on a specific map.
To go even further, it may happen that rendering only one of those values induces mappers to misuse this tag as a "label".

There may be public transport operators that aren't part of any network, or where the network isn't tagged by a mapper because its value is identical to the operator.

So, if the _feature_ we want to show is "for what can you buy tickets here?", I think showing the value of network as default and the value of operator as fallback if no network tag is given is the way this can be achieved.

[DerDings]

I gathered some more examples:

France:

• Lille

Hungary

• Szeged