[Talk-us] Blue Ridge Parkway

OSM Volunteer stevea steveaOSM at softworkers.com
Tue Jan 31 01:44:01 UTC 2017

Hello again, Kevin:

On Jan 30, 2017, at 5:00 PM, Kevin Kenny <kevin.b.kenny at gmail.com> wrote:
> The use of UK English as the OSM standard gets in the way
(of boundary=national_park)
> here, since
> the UK has nothing that's a direct parallel to a state park. States of
> the US at least hypothetically share sovereignty with the Federal
> government; state parks are the creation of a sovereign entity.

States in the USA ARE sovereign.  VERY briefly (as this is not a political forum), we have "two distinct and separate sovereignties within the same territorial space, each of them restricted in its powers and each within its sphere of action prescribed by the Constitution of the United States, independent of the other."  (Ableman v. Booth, 21 How. 523, US Supreme Court).  As a specific example, ""The United States and the State of California are two separate sovereignties, each dominant in its own sphere."  (Redding v. Los Angeles (1947), 81 C.A.2d 888, 185 P.2d 430).  The constitution, along with the 10th amendment, specify the "enumerated powers" of the federal government, in just a few pages.  It is a remarkable political system.  The bottom line is that there are 51 separate sovereign entities encompassed by the USA, one of them "federal."

While I fully respect and agree with your perspective about the Adirondacks and Catskills, state sovereignty isn't especially germane to this discussion, so I attempt to hew to OSM and tagging "standards."

> In fact, the Adirondack and Catskill Parks in New York State (which,
> incidentally are not State Parks, but rather constitute the State
> Forest Preserve) enjoy stronger legal protection than any of our
> National Parks.

Personally, I have no problem whatsoever with tagging such entities in New York State as boundary=national_park (and/or boundary:type=protected_area + protect_class=2, the newer way to say the same thing, even as it takes the older legacy method to render).  My reasoning is simple and two-fold:  New York State is sovereign, and these "park-like" entities, as you say, rise to at least (if not above) the same level of protection as that established by a national_park.  In other words, when the semantics "mean" what they do, it is correct to apply the syntax which denotes them, subtle political distinctions (as we have in the USA) shouldn't confuse us.  However, let's do allow "sovereignty" to reign as a determining authority.

> I know that when I worked on retagging the units of the Forest
> Preserve, I also wound up having some conversations about
> protect_class. Some people here were reluctant to assign
> protect_class=1b to the Wilderness, Wlld Forest, Canoe and Primitive
> areas of these facilities, because they are not Federally designated
> wilderness. My understanding is that IUCN does not distinguish
> protected areas based on what government extended the protection.
> These are all administered as wilderness (in various degrees), and 1b
> is a better fit than any other designation.

We continue to agree here.  National and state in the USA have "equivalent" sovereignty, see above (though, by convention, admin_levels of 2 and 4 might seem to put one subordinate to the other, which isn't necessarily true).  Therefore, something with a semantic of "wilderness" gets 1b, whether "state" or "federal," as both are sovereign and can make such declarations authoritatively and therefore create the effect of what OSM volunteers experience "on the ground" when it comes to "what do I tag for this land?"

> The legal designation of these lands is complex, and doesn't really
> fit any other system. I made some arbitrary decisions with assigning
> IUCN classes - and then left NYSDEC:category and NYSDEC:class fields
> in place that carry in words the legal designations in case they
> should be of interest in the future. Most importantly, in my opinion,
> I documented all the decisions at
> http://wiki.openstreetmap.org/wiki/NYS_DEC_Lands and
> http://wiki.openstreetmap.org/wiki/Import:_NYCDEP_Watershed_Recreation_Areas

Thank you for your dedicated work here!

> I'm really fine with less-precise tagging for rendering: few renderers
> will care about the subtle legal difference between a Wild Forest and
> a Forest Preserve Detached Parcel, or a Shoreline Protection Area and
> a Tidal Wetland. The detailed legalities are there in the tagging for
> those that do care, and the tag values are documented.

Which is why it is true that while a renderer might not show you such subtle distinctions, if it really matters, one can "dig down to the data" within the OSM database, and because we have established tagging patterns (which evolve and improve) AND IF a conscientious data entry volunteer tagged "properly and completely," there you will find the answer.  That's enough, and resonates well and true with OSM tenets.

> The protect_class really should pair with a protection_object, and
> choosing the protection_object generally winds up informing the choice
> of protect_class. For instance, the New York City-owned recreation
> areas in the Catskills, while they are popular recreational
> destinations for hiking, hunting, fishing, trapping, boating and bird
> watching, all have as their principal objective the protection of the
> area from development to assure water quality in New York's
> reservoirs. That fact informed 'protection_object=water', which in
> turn guided the somewhat unusual choice of 'protect_class=12'.  State
> Historic Sites got 'protection_object=heritage' (and a stack of
> heritage=* tagging), which in turn informed 'protect_class=22'. The
> National Park Service-owned corridor of the Appalachian National
> Scenic Trail got protect_object=recreation, which implied
> protect_class=21. And so on, and on, and on, through dozens of
> different special cases.

Wow, I obviously need to do much more research here, I'm lagging in my knowledge, thank you for the history lesson and pointers.

> I used leisure=nature_reserve for a great many of these areas. (Nature
> reserves encompass a wide variety of things.) At such time as
> protected_area is better supported, the protect_class will give better
> information, but in the meantime, it gives us something to render.
> That is important. People relate to parks, and expect to see them on
> maps. The fact that the rendering has not caught up with the precision
> of the data model is no excuse to make them disappear from the maps.

Agreed, though this is a fuzzy line.  We might start another side discussion (something I've wanted to see for over 7 years) of how our rendering pipeline improves.  It is slow, but sure, yet it is entirely opaque to the average OSMer.

> In short, I tag (in a very limited and specific way) for the renderer
> today AND for the data model tomorrow. I need both the detailed
> information of how the land is administered and to have the parks show
> up on today's maps. I'll be happy to remove inappropriate
> national_park and nature_reserve once the renderers offer me something
> better to call the things.

I think an excellent argument can be (just was!) made to do this, especially as you include "in a very limited and specific way."

> As an exception, I'd actually retain the national_park designation for
> the Adirondack and Catskill parks. These entities are massive: the
> Adirondack Park is larger than Slovenia. These entities also are
> administered as complex public-private partnerships with a great many
> different classes of protection within them. They are not monolithic
> protected_areas. They also are highly visible to the public, with
> large signs wherever a highway enters and leaves them. And they are,
> in effect, the local government for tens of thousands of people who
> live on inholdings within the parks - surely more visible to the
> locals than their county governments or the hypothetical governments
> of their unincorporated townships.

No argument from me:  to repeat, sovereign states can designate wilderness and similar entities (whether nominally "national" or not).


More information about the Talk-us mailing list