On Thu, Jan 09, 2020 at 11:04:21PM +0000, marc marc wrote:
> Le 06.01.20 à 08:47, Florian Lohoff a écrit :
> > If you have HUGE Buildings i use a node with an address.
> it's amazing the difference in usage.
> I find that addr nodes are very problematic for hudge buildings like
> shopping malls or train stations. the localisation of the node forces
> the routing to go to a specific location when you may be closer to
> another entrance.
> By putting the address on the building, you can not only allow quality
> reverse geocoding for the whole building, but also allow advanced
> routing to select the closest entry instead of going to the preferred
> entry of the contributor who entered the address.

That is a Huge problem putting informations POI, Addresses on Areas. I
addressed this on tagging a while back and was shot down that i should
tag sufficient. But there are tons of problems with non-explicit route

My solution/proposal was to have a route/nav-aid relation. If you want
to reach Object A, with transport method B, you need to navigate to node C.

Just to get an impression about the impact i am creating to 2 debug
views (for myself and other interested partys) for parts of Germany.

A problem in the countryside are long driveways, sometimes not mapped,
tagged with random access tags or falsely tagged as tracks:


OTOH in the dense urban areas you have the problem of Address for road A
nearer to Road B. So you get navigated to the wrong spot on the road
network. This view is generated with the OSRM Car profile and mapping
all addr:* objects with the "nearest" function and comparing the highway
name and the addr:street. If both are "filled" and non equal -> fail.


Just pan around and you'll easily spot horrible problems. The simple
case its just at a corner. But there are a lot of cases where you end up
significantly far away from the real address on a completely unrealated

So yes - there are serious issues with the way OSM maps POIs and
Addresses and their usage for navigation.

Florian Lohoff                                                 f...@zz.de
        UTF-8 Test: The 🐈 ran after a 🐁, but the 🐁 ran away

Attachment: signature.asc
Description: PGP signature

Tagging mailing list

Reply via email to