> But on the other hand, it's a place, not only an abandoned village

A `place=village` a place, though?

Life-cycle tags don't change the semantics of that, afaik?

> But on the other hand, it's a place, not only an abandoned village

I disagree here. It's specifically the name of a village that is now abandoned. There is nothing else there, and if there is in future that place can be added.

> place=locality#Don't_use_place=locality_just_to_force_a_name_to_appear_on_a_feature.

Doesn't that support the current tagging? If we want this to show on the default renderer, we should figure out a tagging combination that makes sense and is accurate to the reality of it being an abandoned village site, but which openstreetmap.org can show.

A further suggestion: these *could* be tagged with `historic=ruins`

I don't *think* this (and most others) can be marked as `heritage=` as they mostly aren't actively marked as suck by one of the Canadian heritage orgs, or included in the sites maintained by Haida Gwaii Watchmen? If there is a link to a heritage org/program, though, then that tag could also be added.

Can you provide an example where `place=locality` exists alongside a more specific `place=` tag?

I'm also not sure what adding `place=locality` would improve here. The name is still there, it is accurately mapped as an abandoned village.

Is your concern that it doesn't show up on the default renderer?

> That bad practice exist(ed) in US

That article says "especially in", not that it was specifically in the US, nor is that the issue it is highlighting. The point that article appears to be making is that `place=locality` is not specific, with e.g. `abandon:place=village` being preferred.

See: abandoned:place=*

It specifically calls this out: "Many such features are currently tagged place=locality, especially in the United States where many GNIS features were imported as place=locality. However, place=locality is a much less specific tag, used for many other features, including places that were never populated."

It's an abandoned village, so the specific tag (recommended for place=locality) would be "village". It's abandoned, so I added the lifecycle tag to note that.

The name is still there ("Chaatl"), of course. Some of the other abandoned villages I added in this and adjacent changesets have geographic features (e.g. bays or other waterways) that share their name, and those are indeed noted.

I *think* that `abandoned:place=village` is the correct tag here?

Ah, probably right! I've modified the ways. (Also: feel free to correct things you see ..)

I marked areas that are paved with sett. Areas marked with concrete were e.g. stairs and one area of tunnel. I'm not sure which path you are referring to in this changeset, and apologies if I marked a path as concrete that wasn't (or isn't now?), but there's clearly a mix of set / stones / concrete in this change set.

Cheers!

Very nice catch on the incorrectly named street!

BTW: Is adding the postal code, town, and / or province to buildings really worth it? This can be found very easily with a spatial search ("building osm.wiki/Tag:addr:housenumber=... osm.wiki/Tag:addr:street=... IN (boundary osm.wiki/Tag:name=...") and prevents having each and every building tagged with what is ultimately redundant data?

I tend to see the spatial relationships ("building XYZ in boundary type X name Z") as being .. well .. relational :)

Oh, and I double checked against Miami-Dade's open map data as well.

The addresses on the houses, which I then re-confirmed with some online searches for those same addresses.

The comment should be *Centredale* of course. :/ Centreville is elsewhere in the province.

Thanks! That was a search in JOSM that went wrong. Thought I'd cleaned it all up, but missed this multi-polygon.