[Talk-us] railway=abandoned and mapping things that are not there any more?

[Nathan Edgars II]

On 7/13/2012 11:56 PM, Peter Dobratz wrote:
> I guess the main thing that is problematic about former railways is
> their verifiability.
>
> Between TIGER and GNIS and other imports, when I come across a road or
> church that no longer exists (or in some cases never existed), I just
> delete them.  But before I delete the objects, I try to make some effort
> to move them to their correct location where applicable.  How am I
> supposed to tell if these railways are in the right spot?  In this
> specific case, I was able to make an argument that the railway was in
> the wrong spot and you managed to dig up a map that's more than 100
> years old to help find the correct former railroad location.

Old USGS topos are a good source for most abandoned railways, and 
they're now mostly online: http://nationalmap.gov/historical/ Even 
modern topos (available in JOSM - "MSR Maps Topo" in the imagery 
preferences) show many that are not obvious on aerials.

> Where does the information about the geometry of these former railways
> come from?  What steps did you take to verify the data before putting it
> in the OSM database?  How are you going to document this information for
> future mappers?  It would seem to make sense to add additional
> information to a source= or a note= tag indicating where the information
> about the previous existence of the railway came from.  Or maybe there
> can be more verbose information added to the wiki somewhere.

Given that most roads added by mappers are at most tagged source=Yahoo 
or Bing, despite these only being used for location and not name or 
other characteristics, sourcing is not in practice a major concern for 
most mappers. However, I understand the point about being difficult to 
verify, and will try to provide better source tags where applicable.