I’m just wondering what is the popular opinion on country names. For example, should one name New Spain as New Spain or as the Viceroyalty of New Spain. I often name countries with their short form name and then include official_name=*. Other countries went through lots of revolutions, so I name them Empire of Mexico, then Republic of Mexico, then Kingdom of Mexico, and then if it’s the current relation, then I just name it Mexico.
This reminds me of a similar issue with building names: in the present day, a historical building may be known by its longest-standing or most famous occupant. I make sure the chronology relation has this historiographical name, then name each member by the contemporary name if it has one, so that the name is accurate with respect to the start and end dates.
Sometimes historians name the building’s eras as well. The building’s occupant may not have become famous until moving elsewhere, but historians may apply their name to that time period anyways. I do set this as the name=* if it has no other name, but otherwise, I’ve tended to stick this name in alt_name=*, essentially for geocoders but not renderers.
At least “Mexican Empire” was the official name, versus “First Mexican Empire” which is purely historiographical. I don’t think it’s a big problem to set name=* to an official name, as long as the common name is then available as short_name=*. That said, long names tend to collide each other out of a map, resulting in less information density.
It’s good that we can search for “British Hong Kong”, but it’s kind of odd that this name shows up on the rendered map during that time period alongside other contemporary names. As far as I know, no one called it by that name; the official name was just Hong Kong or 香港. Maybe “British Hong Kong” would be more appropriate for the inspector than the map. In that case, we could use a dedicated key like historiographical_name=* or retrospective_name=* that the site could look for. (Any ideas about a less unwieldy key name?)
I like the idea of using the common name in name=* instead of the official one (this is also OSM’s choice), since that is what people who consult the map are more familiar with, and also that of adding the form of government in the name of countries that have changed it.
Regarding retrospective names, although they are not accurate, perhaps they are preferable as they allow users to distinguish the political entity that administered a certain territory just by looking at its name (though some might be more controversial than others, such as calling the eastern part of the Roman Empire the Byzantine Empire).
Perhaps in the future a setting could be added to allow switching between the common name and the official one.
1 Like
“British Hong Kong” etc might be influenced by the Wikidata and Wikipedia titles. I would start with label= similar to what’s suggested before for OSM, for a rendering and application label.
1 Like
The problem with retrospective term is it would be conflated with the meaning of name= , which should be what’s valid in that time period. It immediately fails if someone wants to make a distinction between name= and official_name= for that time period.
As a related example, the historian term of Nationalist China would conflict with name=China for official_name=Republic of China , while the short_name= may be occupied by =ROC for abbreviations. It can’t be added to them.
alt_name= is basically a placeholder for giving up when there are no other better *_name= that fits, or it’s really equally common as the name= somehow. It’s a last resort that should be avoided if possible.
1 Like
label=* would be a decent choice for historiographical names when they conflict with what’s “on the ground”. Unfortunately, it might leave the impression that we’re only interested in hacky renderer hints. That’s been a longstanding problem in OSM with the label role for boundary relations.
Another issue, specific to OHM, is that some mappers and imports have historically used label=yes in large numbers (almost 5,000 as of writing) with the apparent expectation that a renderer would somehow prefer these elements’ labels over other elements’ labels in the event of a collision or at lower zoom levels. This isn’t actually how renderers work, so I’d be in favor of deleting these tags en masse to make room for the usage we’re discussing here.
1 Like