New Zealand - Suburbs and Localities

Communities Oceania
,
1

Hi all,
I have a question about LINZ Data Service

Should this data be added to the OSM and if yes what is the preferable way for the community?

My take would be to add the data in a way to create a relations between polygon and place node already present on the OSM. The main issue that I see is if the relation place tag should be the same for all localities or (my preference) to have place value from the corresponding node representing that place?

Any thoughts, suggestions, issues?

2 Likes
2

Hi Aleksandar, thank you for raising this query. I agree it’s an important conversation to have with the community because there may be differing opinions on whether to import this data, and how to maintain the data once imported.

I can certainly see the appeal of having suburbs and localities defined as closed ways/relations, but I also tend to agree with this line from the wiki page for place=suburb:

“Suburbs may have uncertain boundaries, may overlap with other suburbs, and are often best mapped using a node.”

However further in the article it does state:

How to map

  • As a Node: Place a node at the center of the suburb tagged with place=suburb and identify it with a name=*.
  • Alternatively a suburb can be identified with a closed way or relation if the boundary is clearly defined. In this case a type=boundary relation is often used.

The LINZ dataset does give us boundaries. But are these boundaries so clearly defined by the community? Once imported, we would likely see contributors shifting the boundaries per their own subjective experience of suburban boundaries. Would this be accepted, or would we want to consider the LINZ data as authoritative?

My 2cents to get the thread going, cheers :slight_smile:

3

For reference there was a previous discussion of this topic.

1 Like
4

I would use LINZ dataset as starting point. There are no overlaps withing that set. However it may not be accurate per local contributors, but shifting the boundaries in that case would not create an overlap, just more precise boundary. My understanding is that these boundaries may not be so strict, but having them on OMS may bring more benefit than not importing them when we have available resource.

For this other discussion about addresses, we can use place value as it is in the addr:place tag so we can unify the data.

1 Like