[OSM-talk] Post code areas

[Frederik Ramm]

Hi,

Ian Dees wrote:
> Do we really want to import these kinds of administrative boundaries? If 
> a "regular old mapper" can't go out with a GPS and verify that a border 
> actually exists, does this sort of data belong in OSM?

A question that applies to all administrative boundaries, even most 
national boundaries in Europe nowadays. Like you, I am skeptical about 
them and would prefer them being kept in a separate database.

Post code areas are tremendously useful in Germany because they are 
commonly used as a cheap machine readable form of location descriptor 
("enter your post code to find the nearest band branch" etc).

That alone doesn't justify importing them. My main reasoning is

1. there is no free data set of municipal boundaries in Germany
2. therefore I'd like to use OSM to crowd-source that data
3. postcode boundaries will often run alongside municipal boundaries
4. so importing post codes is a good start to achive 2.

In our case, in addition to the above, the only free post code dataset 
available is a bit aged, and unmaintained, and will need to be corrected 
by the crowd.

> It seems like someone that wants to use the data should go get it from 
> the authoritative source and overlay it  on top of OSM "what's on the 
> ground" data.

Generally that would be my idea too if (a) the data is free, (b) being 
well maintained by a third party and (c) not all that useful for us to 
derive anything from. If these conditions are all met then it makes very 
little sense to import the data.

In our postcode case, only (a) is true and (b) and (c) are false, that's 
what lets me want to make an exception.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00'09" E008°23'33"