A large thank-you to Kevin for that deeply informative post.
> brad <[email protected]> wrote:
> I think its simpler and better to just create an inner boundary as was done
> with the Coconino NF
The Coconio NF (relation/10956348) hasn't "an" inner boundary, it has hundreds
of them. I'm not sure I understand what Brad is saying is "simpler and better"
here, as a well-constructed multipolygon in OSM is "a well-constructed
multipolygon in OSM." We already know how to do that so I don't think we want
to develop something else to represent the same thing.
Is Brad or Mike proposing something else, like two multipolygons to describe
one national forest? I'd be against that (unless I hear and understand more).
Perhaps Brad can tell the list what he thinks is "simpler and better" in the
context of a well-defined multipolygon with one or more outer members, one or
more inner members (as we've established) and then what he might propose?
There is something intriguing with how Mike worded it ("separate polygons for
inholdings, tagged with access=private and possibly ownership=private") which
is certainly novel, and I'm willing to listen to that, but I don't quite
understand what he means. Two relations for one forest? Our wider tagging
practices don't (currently) understand this (two relations, one entity), nor do
any renderers (that I know of), but this sort of access/ownership tagging on a
separate polygon is an idea that might allow us to pack semantics into a
relation (or two relations?) in a way I haven't thought of before. Kind of
pie-in-the-sky, but I'll listen, provided I fully understand what is being
proposed.
We've established there are "more simply described" national forests where
more-or-less "only" (or substantially only) the outer polygon is a member of
the relation, and "very well described" national forests with highly complex
memberships (perhaps multiple outer polygons, and numbering into the hundreds
of inner elements, like Coconio). OSM (in my opinion) has room to accept both,
knowing that while the latter is much more complete, the former might be either
a case of "very few if any inholdings, so essentially 'done'" or it might be "a
rough sketch of (only) the outer polygon member to get the relation started,
more inner polygon memberships need to be added to this relation." And that's
OK, but if / as we do so, let's make note of it (perhaps a FIXME tag in the
relation with value "Incomplete; needs more inner members to describe the full
gamut of all inholdings in this forest.")
SteveA
_______________________________________________
Talk-us mailing list
[email protected]
https://lists.openstreetmap.org/listinfo/talk-us
