Nov 14, 2020, 12:34 by [email protected]:

>
>
>> I am not sure which kind of "reduction"  you are speaking about, if it could 
>> make the difference of a whole village being included or excluded I am sure 
>> that it is not an acceptable level of reduction/tolerance. I would rather 
>> see it the opposite way: not so unlikely that the official data has already 
>> been reduced compared to the (usually very precise "internal" official  
>> data), before publishing it as open data, and there might be some resulting 
>> details which could be seen as problematic on the micro level (e.g. if a 
>> road or a waterway is part of the area or not, or maybe is after the 
>> simplification half in half out, etc.) Also these areas will be delimited by 
>> other features (roads, fields, settlements, waterways, etc.), and ideally 
>> there boundaries in OSM should match with the representation of these things 
>> in OSM (this means more consistency as matching the exact same coordinates, 
>> which will be relating to official data. Hopefully these differences are 
>> small anyway, but a few meters can already make the difference whether a 
>> road or a stream is included or excluded, or is half in half out.
>>
>  
> I meant to manually simplify areas. But as you say, I think that's not 
> acceptable: we need to import those areas as they are in the shapefiles. We 
> are speaking about a official dataset that is constantly used in local 
> administration to decide anything (ie. if a house, road... is inside a 
> natural park or not and if requires, can be granted something...).
>
Substantial changes are unwanted. But note that some shapefiles have unneded 
extra nodes,
I have seen simple rectangle building mapped with 1200 nodes where 4 would 
describe the same
shape.

Such simplification are desirable and should be done also here if needed and 
possible without
any real data loss.

_______________________________________________
Imports mailing list
[email protected]
https://lists.openstreetmap.org/listinfo/imports

Reply via email to