Multiple delimited names in the name tag

[Minh_Nguyen (Minh Nguyễn)]

@Matija_Nalis

I mean, local government must have already made a choice, did they not? There are street signs on the ground, right? And on them is some text in some order - that is the text that IMHO should go in name=* tag.

I think this is taking signs too literally. Can a ⚜ or be a delimiter?

File:Signalisation routière bilingue à l'entrée de la Louisiane.jpg

In the bilingual places I’m familiar with, a delimiter between the names in two languages is just a matter of typographical style or signmakers’ convenience. You’ll encounter inconsistency from one sign to another. If you call up the local government and ask which one is official, they’ll roll their eyes at you and hang up. They’ve got better things to worry about.

@Matija_Nalis

(ok mostly kidding about this last one - we should not use newline character in name, but replace it with some common alternative like name=Brussel / Bruxelles)

We should distinguish between the raw data in OSM and how a data consumer will output something to end users. Indeed, the OSM data model doesn’t support newline characters, and newlines would be problematic inline (when displaying a fully-qualified location name inside a list of search results, or in the title of a webpage), but in many cases, there’s nothing inherently wrong about a renderer displaying the multiple names on different lines if there’s enough space.

There are indeed places where the delimiter is specified officially. In New Zealand, a slash officially and commonly separates the English and Māori names of places such as Aoraki / Mount Cook. But arguably the full name is now “Aoraki / Mount Cook”; many of the name:* tags on this node include a slash.

I think this discussion is about less formal cases where people continue to use one name or the other and the delimiter is not something to get hung up on. Why not let the data consumer figure out the appropriate delimiter based on the local language or an intuitive one based on the user’s preferred language? (After all, in some bilingual regions like Hong Kong, the two languages use different Unicode characters for a slash and different characters for a dash.)

Meanwhile, there’s the use case mentioned above, which I think is underappreciated. If a map intentionally displays the preferred-language name followed by the local-language names, for the benefit of the user, then it needs to avoid repeating a name that both the preferred language and one of the local languages happens to share. This is difficult and unreliable if the delimiter is unpredictable.