[Talk-us-massachusetts] address observations from the field

[Greg Troxel]

I had a spare 30 minutes in downtown Maynard yesterday, and made lots of
notes in osmand about map/reality differences.

There were several cases where there was an address on a building, and
also amenity/shop tags, but there were actually two things in the
building.  I separated some to nodes, since it seemed that the address
is of a sub-building part, not the building, same as the amenity/shop.
There were also cases where a business had an address tag of e.g. 60-62
but I think that's the lot and there is a 60 and a 62 (need to check
more).

One building has two businesses, with different addresses, and the
addresses are on different streets (corner, and makes sense really).

So, in processing addresses, that makes me think:

  I'm even more suspicious of ranges than before.  My working assumption
  is now that addresses of the form N-M are always wrong.

  We have talked about matching address points to buildings and adding
  them there.  But that's only ok if there is a single address point
  coordinates that maps to a building.  If more coordinates within, they
  probably should be nodes.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 162 bytes
Desc: not available
URL: <http://lists.openstreetmap.org/pipermail/talk-us-massachusetts/attachments/20180912/4ce725b4/attachment.sig>