[Imports] Importing addresses in New South Wales, Australia

[Dion Moult]

The problem occurs when there are many houses / apartment within a single property boundary. I agree with the following:

1) don't merge the addresses. A mapper with local knowledge would be able to split them up and assign them to different buildings. Merging the addresses would intentionally create incorrect data.

2) don't leave the nodes in the same position. This action has the effect of intentionally creating messy data for humans. (it would be good for navigator software, though! But OSM is humans first!)

3) don't delete the data. It's useful data that has been extracted and shouldn't be thrown away.

My conclusion to satisfy the above is to take those nodes and put it in another separate OSM file, maybe grouped by suburb and called EPPING-UNRESOLVED.osm. I will upload this to the git repository but not upload to OSM.

This way OSM always has clean data, and we keep a record of all of these uncertainties. Others in the future can check them out and resolve them through local knowledge mapping. I reckon after collecting a bunch, they can be resolved through a Sunday afternoon drive around a few suburbs taking notes.

What do you reckon?

Dion Moult

-------- Original Message --------
On 7 Jul 2018, 21:33, Andrew Harvey wrote:

> On 7 July 2018 at 21:18, michael spreng <talk at osm.datendelphin.net> wrote:
>
>> Hi
>>
>> If the addresses just differ in house number you can merge them and put
>> both numbers in the house number tag, separated by comma: 4,6 for example.
>
> I don't think that's the best idea, since they are different addresses.
>
> I do think it's preferable to avoid overlapping points as it makes it more difficult for mappers who come by later to fix things up based on a survey.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/imports/attachments/20180708/18c5828c/attachment-0001.html>