[Imports] Handling of existing buildings in NYC import (was Re: Buildings & Address in Washington, DC, USA.)

[steggink at steggink.org]

Quoting Simone Cortesi <simone at cortesi.com>:

> On Fri, Jun 6, 2014 at 8:39 AM, Katie Filbert <filbertk at gmail.com> wrote:
>> Some of the problematic things I found earlier this week when surveying:
>>
>> http://www.openstreetmap.org/note/179039
>>
>> http://www.openstreetmap.org/note/178197
>>
>> It's difficult to go back and look at the deleted buildings and say which
>> were better, but I don't think any were kept as far as i can tell.  We even
>> put in addresses! and they were deleted. I don't think that was necessary.
>
> It is easy to do such things if you take advantage of non local
> mappers taking decisions based only on satellite imagery and not local
> knowledge.
>
> @alex: how many "import guys" were local from NYC compared to non-local?
>
> --
> -S

It really depends on the data and the satellite imagery. The data  
might be current, but most likely the satellite imagery is not. That  
way a mapper who is not on the actual site cannot make informed  
decisions.

Regarding conflating data (keeping existing buildings which are  
well-surveyed): it might be hard to do if conflicts occur, for example  
when buildings overlap. Something needs to be adjusted. This depends  
on the quality of both the existing and the imported data.

One thing which is neither hard nor difficult is to contact the  
mappers who have surveyed them, and discuss with them what should be  
done with their data. An import should not behave like a steamroller,  
since that would only make the case for future imports more difficult,  
which is completely understandable

Frank