[Imports] Address Data Import Norway

Ruben Undheim ruben.undheim at gmail.com
Wed Oct 1 22:48:55 UTC 2014


Hi Johan,

Thanks a lot for showing interest!

> 1. I agree with Paul on the general view not to add addr:country, since
> country borders are easily maintainable

Good, we will not add addr:country. I agree too :) The wiki-page will
be updated.

> 2. It should be possible to have a single node combining POI info and
> address info, I'm not sure if that can be the case in Norway

When I say POIs, I mean nodes with either amenity, shop, name etc. The
plan is to let such POIs with address information remain in peace, and
then add nodes (or buildings) with only address information. At many
places there are several POIs on one single address. Having one and
exactly one node (or way) with only address information will simplify
maintenance and provide users of common rendering engines with one
visible number on the map.

> 3. Why are you not importing buildings along with the addresses?

I'm not sure if I understand your question. There are no buildings in
this dataset from Statens Kartverk. If we want them, we will have to
draw them, and then we won't be done with the import in a very long
time - considering the number of houses per person here. :) We are of
course always trying to draw new buildings though..

> 4. You write that 'All addresses should exist as separate nodes'. In OSM
> it's common either to attach addresses/poi's to single building outlines (if
> the complete building with all floors has one address/poi) or to have them
> as separate nodes. Depending on local communities imports can either choose
> to attach nodes to building outlines or to import them as single nodes, so
> it's fine that the Norwegian community chooses to import them as separate
> nodes. But why is it not possible √°fter the import to have the common OSM
> practice which allows addresses being attached to buildings?

The addresses in the dataset are nodes. Therefore we prefer to import
them directly as they are with the 4 addr.*-tags. Whenever there
already exists a building in OSM with the same address, the node for
this address will not be imported. Some people in the Norwegian
community wanted to do it like in Denmark - having only addresses on
nodes, but others opposed this. We therefore now aim for a solution
where both are allowed - but the new imported adresses will be nodes.
The web-service that is continuously comparing the data with the
dataset will accept both
(http://osm.beebeetle.com/addrnodeimportstatus.php).

> 5. What is the procedure on abbreviations? (ps tip: in the Netherlands we
> prevent abbreviations in street names so we write them in full; this causes
> a misalignment when updating addresses from the official country database,
> which can be overridden when you compare using the combination of postcode
> and housenumber only)

Luckily, the dataset contains fully written street names! This is
therefore not a problem for us. It has been discovered some typos in
the dataset. For this we have a correction list in xml-format:
https://github.com/rubund/addrnodeimport/blob/master/xml/corrections.xml

> 6. I can recommend Geofabrik inspector as a QA tool to align names of
> streets to streetnames in addresnodes:
> http://tools.geofabrik.de/osmi/?view=addresses&lon=10.75140&lat=59.91387&zoom=14&overlays=street_not_found

Thanks for the tip. Several of us already know about this tool, and we
are using it to add missing roads and names. It's great!


Best regards,
Ruben



More information about the Imports mailing list