[Talk-GB] Pavements (footways/sidewalks) mapped as pedestrian areas AND footways

[Andy Townsend]

On 26/12/2022 13:25, David Woolley wrote:
>
> My impression is that OSM generally has a problem with changing 
> abstractions at different zoom levels.

I don't think that that is the problem here - what's happening here is 
that someone mapping the same thing multiple times as different features.


> I don't know how OS handle this in their map databases, but the 
> problem here is that the user is trying to get the plan type view that 
> appears on the sort of OS maps used for planning applications, to 
> coexist with a much more route planning type abstraction.

I suspect that the mapper here isn't seeing "the OSM data" as distinct 
from "the view that they see of it in OSM Carto".  That's not surprising 
since "a different view of the data" in e.g. Google Maps simply isn't a 
thing.


>
> It seems to be common to map footways over pedestrian areas, even 
> though they don't exist on the ground. 

I suspect it depends where you are.  A quick look locally finds only a 
few, https://overpass-turbo.eu/s/1pqi which are mostly pretty much OK, 
to be honest.

Further out there is https://www.openstreetmap.org/way/1004137860 which 
is a bit iffy (the linear way is 
https://www.openstreetmap.org/way/130352229 ), though knowing the area I 
can understand why the pedestrian area got mapped as it did.


> I've given up arguing on these, as they will reappear if you remove 
> them.  In my view, they are really mapping for the renderer; the 
> argument is that, without them routers either don't route at all, or 
> route around the edge of the area.
>
If the linear way is mapped then surely routers will use that, however 
the area is mapped?

Best Regards,

Andy