[Tagging] Feature Proposal - RFC - boundary=forest(_compartment) relations

[Brian M. Sperlongano]

There is something I think that is missing from this discussion, and that
is the distinction between a *boundary* and an *area*.  This distinction
has no basis other than my own opinion, but here goes.

Something tagged as an "area" means that "the entire internal contents of
this polygon has this meaning".  Thus, when we tag natural=beach it means
"this whole area is a beach" and when we tag an area "landuse=vineyard" we
mean to say that the entire area is used for growing grapes.  In the latter
case, the landuse polygon may happen to coincide with the legal property
boundary of the vineyard, but the area covered by the polygon is meant to
cover the actual ground "used for grape growing" and not the extent of the
vinter's land ownership.

Meanwhile, a "boundary" is not an area, but a perimeter.  The distinction
is subtle but important.  Fundamentally, a boundary has *legal* meaning,
and indicates that "there is a human-constructed and (often) non-observable
institutional difference demarcated by this line".  The institutional
difference might be administrative, political, legal, etc.  Unique amongst
OSM tagging, boundaries are not (necessarily) observable on the ground.
Implicit in a boundary is the idea that they are independent of land use,
land cover, or any other feature contained within.  A boundary says nothing
about what is contained within; it only describes the human institutional
difference that occurs on either side of the perimeter.

I am challenged to understand exactly which of these two categories are
being described here.

If we are describing "land which is used for forestry" then we must
conclude that we are describing landuse, and thus we must turn to the
much-maligned landuse=forest or perhaps some replacement of it.

If we are describing "special forestry laws apply to this parcel of public
land", then we *are* describing a boundary and the burden is on the
proposal to show how this differs from and/or impacts
boundary=protected_area, and how a mapper might interpret and apply that
difference for various types of forestry lands worldwide - for example, a
State Forest in the US or a Quebecois "Zone d'exploitation contrôlée"
controlled harvesting zone.

If we are simply describing "the private land ownership boundary of a piece
of land", then it is cadastre and should not be mapped, consistent with the
community stance on the mapping of private land ownership.

On Wed, Dec 30, 2020 at 1:07 PM David Marchal via Tagging <
tagging at openstreetmap.org> wrote:

> Hello, there.
>
> I designed a proposal for mapping forests and their compartments, by
> separating these features from underlying landuse/natural features, thus
> allowing to link non-wooded areas to their belonging forest :
> https://wiki.openstreetmap.org/wiki/Proposed_features/boundary%3Dforest(_compartment)_relations#Features.2FPages_affected
>
> You may, and are invited to, comment this proposal on its discussion page (
> https://wiki.openstreetmap.org/wiki/Talk:Proposed_features/boundary%3Dforest(_compartment)_relations),
> where a topic already waits your comments for improving the proposal.
>
> Awaiting your contributions,
>
> Regards.
> --
> Sent with ProtonMail <https://protonmail.com> Secure Email.
>
> _______________________________________________
> Tagging mailing list
> Tagging at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/tagging
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/tagging/attachments/20201230/cbac84a6/attachment-0001.htm>