[Talk-us] Federally Funded Research R&D Centers: landuse=military?

[Greg Troxel]

  I, too, am interested in this question.  I have such a facility in
  my neighbourhood, for which I have official cadastral data.
  I asked on IRC about it, and was quite brusquely rebuffed with,
  essentially, "barrier=fence is the only way to map it." (When I
  pointed out that the fence doesn't necessary follow the property
  line, the gist of the reply was, "we don't do cadastral data."

That sounds bizarre to me.  Thoughts:

  There are some not well understood legal issues about mapping and
  military facilities (Internal Security Act of 1950 - read the signs at
  a base some time).  That may cause some people to tell you not to map
  such things in odd ways.  I will only tell you that I find the
  situation murky.

    http://afp.google.com/article/ALeqM5i3BOMCwxbAZg_Nfh9OyIAYPTlSQA

  Certainly OSM shows the boundaries of many military facilities, with
  landuse=military tagging.  Notwithstanding the above, I have heard of
  no trouble from this practice.

  There's a general negative view about loading all parcel boundaries
  into OSM.  Some of this comes from it being a maintenance problem and
  the limitations of the OSM toolset to avoid seeing some kinds of data
  when editing.  Some of it comes from what I find to be a strange
  "personally observable only" culture.  To me the point of the
  verifiable guideline is that someone else could determine if the data
  is correct.  In the case of lot lines, it might take real surveying
  and deed research, but all sorts of people regularly have determined
  where their corners are.  So dismissing lot lines as inherently
  unverifiable doesn't make sense to me.  (I think in time they will be
  included, when tools improve.)

  While importing all parcels has issues, I don't see any issue with
  grabbing single interesting parcels (from an appropriately licensed
  dataset), tagging them, doing some even rudimentary field
  checking/quality control, and importing them.  I've done this for land
  bought by a local land trust, which then became conservation land.
  Interestingly, stone walls from bing and parcel boundaries don't quite
  line up.  But I'm not sure what that means, and someday might find one
  of the handful people in the world who have a better idea than I do of
  where the boundaries are.  In the meantime, the map is better than
  before I added the lot polygons.

So I'd say don't be dissuaded by people on IRC from adding a polygon
from cadastral data.  Those people can then hand survey and if they,
with due understanding of error sources, believe that their data is
better, adjust it.

(I'm going to explicitly decline to give you advice about the legal
issues, because a) IANAL and b) I don't understand them well enough, so
even if I were I wouldn't. :)