[Talk-GB] Automated Code-Point Open postcode editing (simple cases only)
ndrw6 at redhazel.co.uk
Wed Jul 17 10:58:54 UTC 2019
I haven't seen such guidelines for the UK myself. In general people prefer having address tags on buildings. Separate address points are more of a stop gap solution until building shapes are sufficiently accurate.
An exception is addresses on entrances, which some mappers seem to prefer. I can simply avoid such areas.
- Code-Point Open contains only one data entry (point) per postcode. It is located near the centroid and snapped to a nearest building (technically delivery point). So the scope is fairly small.
- This is automatic editing, not an import. I can skip areas that use different addressing conventions. There is still over a hundred thousands of simple cases that can save a lot of manual work.
- I ignore postcodes that are already in OSM near that location. So it is an additive process (no changes or deletions)
- Manual methods are not perfect either, due to Code-Point Open limitations. But it is still the only legal source of postcodes in bulk we have (licensed by the owner).
On 17 July 2019 10:07:52 BST, Mateusz Konieczny <matkoniecz at tutanota.com> wrote:
>16 lip 2019, 23:19 od ndrw6 at redhazel.co.uk:
>> added as separate points rather than tags (automated edit will add
>addr:postcode tags directly to the building, this is what I chose to do
>manually as well)
>Duplicating address data or adding
>part to a separate node and part to
>building outline seems incorrect to be.
>At least in Poland address imports
>are obligated to handle this situation.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Talk-GB