Tagging of State Parks in the US

classic Classic list List threaded Threaded
49 messages Options
123
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

dieterdreist


sent from a phone

On 28. Jul 2019, at 22:23, Kevin Kenny <[hidden email]> wrote:

For specific kind of sites (e.g. protected under a specific international treaty) we could have specific tags to identify them if desired, e.g. protection_context=natura2000
or
protection_context=state_park
(not sure the latter would be adding information if there was already an admin_level=4 tag)

Already there - the 'protection_title' tag; plus, as you note,
'related_law'.It's pretty useless for rendering, though; there are too
many administrative jurisdictions with too many detailed differences
in their local laws.


this is not the same as protection title. It would make sense for cases like natura2000, i.e. protected sites that are also listed in the natura 2000 network.

related_law is ok for all sites designated by law, while it might also be seen to extend to international treaties, it would probably not be suitable to tie together networks of sites that are controlled by non-governmental actors.


Cheers Martin 

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

dieterdreist
In reply to this post by Kevin Kenny-3


sent from a phone

> On 28. Jul 2019, at 22:23, Kevin Kenny <[hidden email]> wrote:
>
> But this doesn't really address the problem. We can't fix State Parks
> by making them 'boundary=national_park admin_level=4' because they
> don't function as 'national park' in the IUCN deffinition of the term.


the proposal was boundary=protected_area
admin_level=4
you could also add a protection title.

Cheers Martin

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

Paul Allen
In reply to this post by Kevin Kenny-3
On Sun, 28 Jul 2019 at 22:35, Kevin Kenny <[hidden email]> wrote:

Nevertheless, to some extent, we're dealing with "the language of
OpenStreetMap is UK English as interpreted by Germans,"

Sounds like the setup for a joke.  Or a goat song.

Would it be appropriate to propose a mechanical edit to add area=yes
to closed ways that are tagged boundary={aboriginal_lands,
national_park, protected_area} and lack any other keys that would make
them polygons?

Wearing my pedant hat, I'd say it's appropriate to propose just about anything, however
nonsensical.  Taking off my pedant hat, I'd say that seems like a sensible thing to do.
Putting on my cynic hat, I'd say you'll probably get too many objections for it to happen:
people will say you have to manually ensure area=yes is actually valid in each situation;
others will say that mechanical edits should never be performed for any reason; and
a few will say we should not compensate for database/toolchain shortcomings by
adding unnecessary tags.

--
Paul


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

Kevin Kenny-3
In reply to this post by Paul Allen
On Sun, Jul 28, 2019 at 5:42 PM Paul Allen <[hidden email]> wrote:

> On Sun, 28 Jul 2019 at 21:25, Kevin Kenny <[hidden email]> wrote:
>
>> But this doesn't really address the problem. We can't fix State Parks
>> by making them 'boundary=national_park admin_level=4' because they
>> don't function as 'national park' in the IUCN deffinition of the term.
>> Instead, the typical State Park is a hybrid of nature_reserve and
>> recreation_ground and park and maybe a few other things. Requiring
>> that those land uses be mapped separately leaves no whole to which the
>> name and boundary can be assigned, but the whole doesn't really have
>> anything binding it together other than a protection status, a
>> coterminous set of boundaries and a name.
>
> Doesn't seem to fit national parks in the UK either.  See Pembrokeshire Coast National
> Park: https://www.openstreetmap.org/relation/165598
>
> It is entirely within, and occupies a large part of, the county of Pembrokeshire.  It is not
> administered by the UK Government, the Welsh Assembly or Pembrokeshire County
> Council but by the Pembrokeshire Coast National Park Authority.  See
> https://en.wikipedia.org/wiki/Pembrokeshire_Coast_National_Park
>
> The PCNPA owns less than 2% of the national park, the rest of it is privately owned.  It contains
> 13 Special Areas of Conservation, 5 Special Protection Areas, 1 Marine Conservation Zone,
> 7 National Nature Reserves, 60 Sites of Special Scientific Interest and 265 Scheduled
> Ancient Monuments, all of which come under one or another protection scheme and are
> administered by different organizations.  See
> https://www.pembrokeshirecoast.wales/default.asp?PID=503  There are also woodlands
> administered/protected by various different organizations.
>
> It also contains hamlets, villages, town and cities.  As well as everything else you might expect
> to find in an area with a resident population of around 22,500 such as golf courses, recreational
> parks, etc.
>
> Planning permission within the park is controlled by the PCNPA rather than the County Council.
>
> It's all kind of complicated.

The Adirondack and Catskill Parks in New York are nearly a precise
parallel, including the complication that they are not chartered by
the Federal government. They're entirely within, and a creature of,
the State of New York.  (The Adirondack Park spans twelve counties -
at 24,000 km², it's the size of some European countries. The Catskill
Park is perhaps a sixth its size). They are read into the New York
State Constitution, so the unusual cooperative protection that they
enjoy is actually stronger than that of any of our National Parks
(which can, in theory, be wiped out by a simple act of Congress,
rather than a constitutional amendment).

The land ownership is roughly 50-50 public/private. Many of the
private landowners are logging companies, who are restricted to
sustainable harvest techniques and severely restricted against
repurposing the land. (Many also have public access easements in place
so that some areas not being actively logged are open to recreational
users.)

The local government system is complicated, but in the Adirondacks,
suffice it to say that the Adirondack Park Agency (a private-public
partnership between the local communities and the NY State Department
of Environmental Conservation) is the most powerful among the local
entities. Certainly, listings for property are more likely to say
whether they're inside or outside the park than what township or
county they're in, because that difference is more significant to the
landowner.

For the Adirondack Park alone, there are about forty Wilderness Areas
(comprising about 4800 km²), a somewhat greater number of Wild Forest
[a classification just below Wilderness] (about 5400 km²), about 5900
km² of Resource Management areas (in private hands, and mostly devoted
to logging), about 4000 km² devoted to Rural Use (mostly agriculture).
About 50 km² are even zoned as Industrial Use - mostly mining; the
park produces titanium, garnet and wollastonite.
About 1600 km² are devoted to human habitation, divided into about 60%
low-intensity (< 75 buildings/km²), 30% medium-intensity (75-200
buildings/km²), and 10% high-intensity (where there is no fixed limit
and most uses are permitted, although special hearings are needed for
subdivisions exceeding 100 lots, expansion into wetlands, airport
construction, and structures over 12 m above average terrain). It
contains also Primitive, HIstoric, Canoe, and Intensive Use areas.

There are about 130,000 permanent inhabitants, and the population
expands to about 320,000 seasonally, with nearly 10 million annual
visitors. There are two universities (Paul Smith's College and State
University of New York Environmental Science and Forestry), plus North
Country Community College. (There are a handful more just outside its
boundaries). SUNY-ESF maintains multiple research stations and
demonstration forests.

Some of the 'private' land is actually in the hands of NGO's, but this
situation is commoner in the Catskills, where the Nature Conservancy,
the Open Space Institute, the Catskill Center, the Appalachian
Mountain Club, the YMCA, and numerous other entities all have
substantial holdings. Even some of the private landowners welcome
visitors - I've hiked (with advance permission, which is ordinarily
granted outside certain specific seasons) on one of Jay Gould's
(https://en.wikipedia.org/wiki/Jay_Gould) estates several times.

There are no sizable cities in the park, but dozens of towns and
villages of a few thousand inhabitants each.

 I could get the numbers for the Catskills, but this message is
already too long.

Planning permission in the Adirondack Park is controlled by the
Adirondack Park Agency, not the zoning boards of the individual
townships or counties. Catskill Park zoning is complicated because of
the towering presence of the New York City watershed - so I'm not
discussing it, but it's differently restricted, with New York City
having a say in it despite the fact that none of the park is within
city limits.

You're right that it's 'kind of complicated.'

I tagged the things 'boundary=national_park' and I'm not apologizing.

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

Kevin Kenny-3
In reply to this post by dieterdreist
On Sun, Jul 28, 2019 at 5:47 PM Martin Koppenhoefer
<[hidden email]> wrote:

> > On 28. Jul 2019, at 22:23, Kevin Kenny <[hidden email]> wrote:
> >
> > But this doesn't really address the problem. We can't fix State Parks
> > by making them 'boundary=national_park admin_level=4' because they
> > don't function as 'national park' in the IUCN deffinition of the term.
>
>
> the proposal was boundary=protected_area
> admin_level=4
> you could also add a protection title.

Right. And the original proposal was to activate a protect_class=*
that was previously dormant, because nothing of 1a-6 fits the
situation. Many of these parks simply aren't created to conserve
nature, they're created to provide opportunities for a variety of
outdoor recreations.

admin_level isn't *wrong*, it just doesn't help. With or without
admin_level, there's no IUCN-defined protect_class that fits.

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

Kevin Kenny-3
In reply to this post by Paul Allen
On Sun, Jul 28, 2019 at 5:56 PM Paul Allen <[hidden email]> wrote:
> On Sun, 28 Jul 2019 at 22:35, Kevin Kenny <[hidden email]> wrote:
>> Would it be appropriate to propose a mechanical edit to add area=yes
>> to closed ways that are tagged boundary={aboriginal_lands,
>> national_park, protected_area} and lack any other keys that would make
>> them polygons?
>
> Wearing my pedant hat, I'd say it's appropriate to propose just about anything, however
> nonsensical.

Thanks for the levity!

> Taking off my pedant hat, I'd say that seems like a sensible thing to do.
> Putting on my cynic hat, I'd say you'll probably get too many objections for it to happen:
> people will say you have to manually ensure area=yes is actually valid in each situation;

Yeah. Although, how can a protected_AREA be anything but an area?  But
never mind that!

> others will say that mechanical edits should never be performed for any reason;

Still, they get done from time to time. I've even seen bots come by
and tidy up things in my own work.

> a few will say we should not compensate for database/toolchain shortcomings by
> adding unnecessary tags.

Yeah. The ones who ignore the statement that you just made, " If it
breaks things, it won't be used."

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

Joseph Eisenberg
how can a protected_AREA be anything but an area?”

Right. Please don’t add area=yes to these features. 

This tag is only needed for features that can be either a linear feature OR an area, for example barrier=hedge.

(Mapping large protected areas mapped as closed ways to relations of type=boundary (or type=multipolygon) is also an effective way of specifying that they are areas, though again I would not recommend retagging existing features.)

Joseph

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

dieterdreist
In reply to this post by Kevin Kenny-3


sent from a phone

> On 29. Jul 2019, at 01:17, Kevin Kenny <[hidden email]> wrote:
>
> admin_level isn't *wrong*, it just doesn't help. With or without
> admin_level, there's no IUCN-defined protect_class that fits.


I didn’t know we were bound to IUCN classes. IMHO we can have our own system, while it should ideally allow to distinguish all the IUCN classes, it doesn’t mean we cannot have more qualifiers, if they seem useful.

For the Italian situation, a distinction for many protected areas by national, regional (al 4), provincial (al 6) and municipal (al 8) seems to make a lot of sense. If it doesn’t apply (e.g. not a protected area by the competent government), don’t put admin_level.


Cheers Martin
_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

dieterdreist
In reply to this post by Kevin Kenny-3


sent from a phone

On 29. Jul 2019, at 01:18, Kevin Kenny <[hidden email]> wrote:

Putting on my cynic hat, I'd say you'll probably get too many objections for it to happen:
people will say you have to manually ensure area=yes is actually valid in each situation;

Yeah. Although, how can a protected_AREA be anything but an area? 


+1, I don’t see why we would need area=yes on boundary=protected_area

It also seems weird that a node is suitable according to the wiki:

but this is a general semantic problem we have with all boundaries (a boundary is a linear object, what it delimits is an area, which is somehow implicit through the boundaries, but when we describe the area inside with tags we add these directly to the boundary object, rather than to an area object with the boundaries as limits).

Anyway, area=yes seems superfluous as we are always speaking about areas in this context, even if they are represented as a node.

Cheers Martin 

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

Paul Allen
In reply to this post by Kevin Kenny-3
On Mon, 29 Jul 2019 at 00:08, Kevin Kenny <[hidden email]> wrote:

There are no sizable cities in the park, but dozens of towns and
villages of a few thousand inhabitants each.

I can think of only one city in the Pembrokeshire Coast National Park (but I'm not that
familiar with it) and that's the second-smallest city in the UK by population: St David's
with 1,841 people in 2011).  Note that, in the UK, official city status must be conferred
by Mrs Betty Windsor or her predecessors.  Unofficially, some very large towns
describe themselves as cities without having that status officially.

I tagged the things 'boundary=national_park' and I'm not apologizing.

+1 for the tag.  +1 for not apologizing.

--
Paul


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

Kevin Kenny-3
In reply to this post by Joseph Eisenberg
On Sun, Jul 28, 2019 at 7:58 PM Joseph Eisenberg
<[hidden email]> wrote:
> “how can a protected_AREA be anything but an area?”
>
> Right. Please don’t add area=yes to these features.

If you're saying that the alternative is for mappers to wait, possibly
years, for an approved osm2pgsql stylesheet+Lua change, and a full
database reload, before these features can be rendered, then that's
simply not going to work. Mappers *will* work around that sort of
restriction, and the goal here is to guide the community to the least
damaging work-around.

> This tag is only needed for features that can be either a linear feature OR an area, for example barrier=hedge.

And features that the osm2pgsql importer hasn't learnt are areas.

> (Mapping large protected areas mapped as closed ways to relations of type=boundary (or type=multipolygon) is also an effective way of specifying that they are areas, though again I would not recommend retagging existing features.)

Right. I've done a fair number of single-member route relations for
Waymarked Trails, and single-member multipolygons for these areas is
indeed a workable alternative. area=yes is considerably simpler,
though, and won't leave us with mystery multipolygons to confuse us
later.

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

Kevin Kenny-3
In reply to this post by dieterdreist
On Mon, Jul 29, 2019 at 4:22 AM Martin Koppenhoefer
<[hidden email]> wrote:
> I didn’t know we were bound to IUCN classes. IMHO we can have our own system, while it should ideally allow to distinguish all the IUCN classes, it doesn’t mean we cannot have more qualifiers, if they seem useful.

We return to the original idea proposed at the very start of this
thread: 'protect_class=21 protection_object=recreation' for these
features. Except for the ugliness of using numeric values for
protect_class, it sounds as if you might agree with the original idea?

> For the Italian situation, a distinction for many protected areas by national, regional (al 4), provincial (al 6) and municipal (al 8) seems to make a lot of sense. If it doesn’t apply (e.g. not a protected area by the competent government), don’t put admin_level.

Sure. I have no problem with admin_level. If you want to tag,
admin_level=4 on state parks, be my guest! It's just a little
distracting, because it doesn't actually address the issue (a area
protected for diverse recreational uses, partaking of park,
recreation_ground, nature_reserve, and a few other things) but with a
single enclosing boundary, and a single name.

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

Paul Allen
On Mon, 29 Jul 2019 at 13:52, Kevin Kenny <[hidden email]> wrote:

We return to the original idea proposed at the very start of this
thread: 'protect_class=21 protection_object=recreation' for these
features. Except for the ugliness of using numeric values for
protect_class, it sounds as if you might agree with the original idea?

I don't know if I would or not because I didn't examine it in any detail after seeing the
numeric values.  The numeric values convinced me it was a non-starter, so I didn't
investigate further.

I have no objection to using it as optional, supplemental information about an object tagged
in some other way.  Much the same as opening_hours or the UK Food Hygiene Rating Scheme
ID with fhrd:id=n.  I can add those, but I don't have to.  With that sort of usage I could even
live with numeric values for protect_class (I doubt I'd ever add one because it's too much
like hard work figuring out what number to use, but if I encountered one I'd not remove it).

I have a problem with the numeric values if you intend this to be a top-level tag for an object,
replacing other ways of tagging such an object.  In that situation, numeric values are (for me) a
no-go.

I suspect whatever you come up with may have to be usable as both top-level and supplemental,
to cater for all sorts of existing objects which already have existing schemes but which also
fall under an IUCN class.  In the UK, Areas of Outstanding Natural Beauty, Nature Reserves,
Sites of Special Scientific Interest and Registered Historic Landscapes come to mind.  There
are others.

--
Paul


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

dieterdreist
In reply to this post by Kevin Kenny-3


sent from a phone

> On 29. Jul 2019, at 14:50, Kevin Kenny <[hidden email]> wrote:
>
> On Mon, Jul 29, 2019 at 4:22 AM Martin Koppenhoefer
> <[hidden email]> wrote:
>> I didn’t know we were bound to IUCN classes. IMHO we can have our own system, while it should ideally allow to distinguish all the IUCN classes, it doesn’t mean we cannot have more qualifiers, if they seem useful.
>
> We return to the original idea proposed at the very start of this
> thread: 'protect_class=21 protection_object=recreation' for these
> features. Except for the ugliness of using numeric values for
> protect_class, it sounds as if you might agree with the original idea?


sure


>
>> For the Italian situation, a distinction for many protected areas by national, regional (al 4), provincial (al 6) and municipal (al 8) seems to make a lot of sense. If it doesn’t apply (e.g. not a protected area by the competent government), don’t put admin_level.
>
> Sure. I have no problem with admin_level. If you want to tag,
> admin_level=4 on state parks, be my guest! It's just a little
> distracting, because it doesn't actually address the issue (a area
> protected for diverse recreational uses, partaking of park,
> recreation_ground, nature_reserve, and a few other things) but with a
> single enclosing boundary, and a single name.


Maybe you can see (and map) the state park as one thing and the nature reserve within it as another? For the state park you would need to say it is a state park and has this name and or number (usually there will be an identifier, maybe it’s not strictly necessary). The nature reserve within would already be spatially connected, but if it is an explicit constituent part of the park that would maybe be too weak? Is the nature reserve managed by the state or a different government level?

Cheers Martin
_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

Kevin Kenny-3
On Mon, Jul 29, 2019 at 9:44 AM Martin Koppenhoefer
<[hidden email]> wrote:
> Maybe you can see (and map) the state park as one thing and the nature reserve within it as another? For the state park you would need to say it is a state park and has this name and or number (usually there will be an identifier, maybe it’s not strictly necessary). The nature reserve within would already be spatially connected, but if it is an explicit constituent part of the park that would maybe be too weak? Is the nature reserve managed by the state or a different government level?

Sure, detailed mapping inside the parks is possible and encouraged!
The fundamental problem has been how to tag the whole park - because
that's what the name is bound to, and many of the areas inside are
less formal - the parks department has decided to keep the 'nature
reserve' part in a natural state for hiking and bird watching, but
it's not otherwise a 'nature reserve' with formal protection. They
hypothetically could choose a different management strategy under a
different administration. (In practice, they don't. You can't turn
that ship very fast.)

In an earlier message (I see that you've not been following the whole
thread, which is getting pretty long so I don't blame you) I mentioned
Bear Mountain State Park
https://www.openstreetmap.org/relation/6467468 as being fairly typical
of a 'complex' example, where the larger-scale mapping is starting to
come along. (With many of these parks, we're still doing well just to
get them on the map in the right places!) Most State Parks are
simpler, but I choose this example because it's the sort that appears
to have nearly "one of everything."

It's a complex object; 'park', 'nature_reserve', 'recreation_ground',
'national_park' are all inaccurate. It's best to give it a border, a
name, a protection status, and call it a day. I contend that the
current tagging of https://www.openstreetmap.org/relation/6467468 is
close to 'best current practice'. It's important to emphasize
'current' here - the numeric protection classes are what the Wiki
currently recommands, and what New York currently uses. That's subject
to change, and the other posters in this thread are demanding such
change. (That's fine, but my original objective was to clarify best
*current* practice!)

(Details follow. Stop here if not interested, but a few paragraphs
detailing the sort of complexities encountered are probably in order.)

The park's border is complex, because the village of Fort Montgomery
and the hamlet of Jones Point are 'inside' the park. (Not quite
inside, topologically, and not owned or administered by the park, but
you have to drive on roads with park land on either side to get to
them)

The parts of the park that see the greatest number of visitors are the
developed sections - the northeast area, near the bridge over the
Hudson, and the scenic drive and observation tower on Bear Mountain
itself (the peak west of Hessian Lake). The main developed area
includes two formal historic preserves (Fort Clinton and Fort
Montgomery - note the 'museum' icons at high zoom levels), a zoo,
swimming and boating facilities, a large grassy area for informal
athletic events, a carousel, several playgrounds, two inns (one of
which also offers group accommodations in large cabins), a ranger
station and various other PoI's. It disallows camping, but camping is
available at the larger Harriman State Park, coterminuous to the west.
(The nearest authorized 'back country' campsite is West Mountain, in a
cutout from the park on the southwest side. The nearest 'front
country' camping is at Silver Mine and Lake Tiorati, a short distance
to the southwest on Seven Lakes Drive (Road, Parkway; it's the same
road, and the signage is inconsistent). The park also embeds an
ecological research reserve (Iona Island - on the Hudson River, on the
east side of the rail grade.

There are other historic sites embedded in the park, such as the ghost
town of Doodletown (the last inhabitants were finally forced out in
the 1960's; the cemeteries are still maintained, and the waterworks
still serve Jones Point, Iona Island and Buckberg), the
never-completed Dunderberg Spiral Railway
https://www.openstreetmap.org/relation/2138097 (an insane
nineteenth-century plan to build what would have been the world's
longest roller-coaster right to the present day; the plan was
torpedoed when the Columbian Exposition of 1892 went to Chicago
instead of New York), and many abandoned mines, including a failed
attempt by Thomas Edison at electromagnetic refinement of iron ore.

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

dieterdreist


sent from a phone

> On 29. Jul 2019, at 16:37, Kevin Kenny <[hidden email]> wrote:
>
> There are other historic sites embedded in the park....


are all these sites mentioned to be part of the state park, or do they simply happen to be within the boundaries?
If the definition of the park is a list of areas and sites, maybe a site relation would be appropriate: you can specify a perimeter and also add stuff inside (not only polygons but also nodes and ways). If legislation said the park is made up of A, B, C, D and E, it would be comprehensive if we modeled it as an object with A, B, C, D and E as members.

Cheers Martin
_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

Joseph Eisenberg
-1 to a site relation for an area with a defined outer boundary.

Relation = boundary (and =multipolygon) works fine for defining an area, and you can make holes to exclude at my “outparcels” or villages which are not part of the official protected area.

Mappers don’t need to add things to relations when the geometry is enough to show that node A isi side of area B.

Joseph



On Tue, Jul 30, 2019 at 7:02 AM Martin Koppenhoefer <[hidden email]> wrote:


sent from a phone

> On 29. Jul 2019, at 16:37, Kevin Kenny <[hidden email]> wrote:
>
> There are other historic sites embedded in the park....


are all these sites mentioned to be part of the state park, or do they simply happen to be within the boundaries?
If the definition of the park is a list of areas and sites, maybe a site relation would be appropriate: you can specify a perimeter and also add stuff inside (not only polygons but also nodes and ways). If legislation said the park is made up of A, B, C, D and E, it would be comprehensive if we modeled it as an object with A, B, C, D and E as members.

Cheers Martin
_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

Kevin Kenny-3
In reply to this post by dieterdreist
On Mon, Jul 29, 2019 at 6:02 PM Martin Koppenhoefer
<[hidden email]> wrote:
> > On 29. Jul 2019, at 16:37, Kevin Kenny <[hidden email]> wrote:
> > There are other historic sites embedded in the park....
> are all these sites mentioned to be part of the state park, or do they simply happen to be within the boundaries?

I'm not sure what you mean by the question.

Many parks over here have 'inholdings', which are other land in the
parks not owned or managed by the park administration. I'm consistent
about mapping those, which is why park boundaries are usually
multipolygons.

The sites that I mentioned in Bear Mountain are mostly administered by
the park administration. The inns are operated by a contractor (but
the operator is responsible to the park management), the hiking trails
are maintained by an NGO, and the research reserve is some sort of
cooperative arrangement with NYS Department of Environmental
Conservation and a couple of other government agencies, and I don't
know the all the details. To a visitor, all but the last are 'part of'
the park. (The research reserve has more restricted access, but is
still managed in part by the park under the cooperative agreement.)

The park's boundaries are exactly (well, to within the limits of the
data source, which is not 100% trustworthy, but I verified against
highway, railroad and river alignment, and against a couple of other
data sources for the boundary with West Point) as shown by the
multipolygon https://www.openstreetmap.org/relation/6467468. The
rights-of-way for the railroad and the highways, and both of the
settlements that I mentioned, are not part of the park, and are cut
out of the multipolygon. I didn't need to use inner ways for those,
simply because they all connect to the world outside, but we do have
protected areas with quite complex topologies indeed:
https://www.openstreetmap.org/relation/6365096 is an example and
https://www.openstreetmap.org/relation/6362702 is a worse one that
took many hours to map. even with the aid of an import. (I *don't*
just drop imported data in blindly!)  The Fort Montgomery historic
site is technically a separate site (Fort Clinton *is* part of Bear
Mountain State Park), and is mapped thus - but it, Bear Mountain, and
Harriman and Sterling Forest to the west are under common management.

Yes, one of the inn buildings encroaches on the Palisades Parkway
right-of-way. The agencies that administer those two objects are
friendly.

The concessionaires rent their stalls.

The rest of the historic sites, I've not done the research to see if
they're listed. Many of them are off limits - you're supposed to stay
on trail in the backcountry in that park (because many of the old
industrial sites are still hazardous - they don't want people falling
down mine shafts or being injured by abandoned and decaying
machinery). Doodletown has interpretive signs at a lot of its ruins,
so I wouldn't be surprised to find NRHP listings there, but I haven't
yet looked.

I'm not sure what I'd do with a site relation here. I'd use it, if for
instance, the park had an office occupying part of a building that
wasn't in the park, so I'd need a node for that. I've done site
relations for urban universities that have outreach programs in
off-campus space, but why bother with a site when a multipolygon will
do?

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

dieterdreist
In reply to this post by Joseph Eisenberg


Am Di., 30. Juli 2019 um 00:51 Uhr schrieb Joseph Eisenberg <[hidden email]>:
-1 to a site relation for an area with a defined outer boundary.

Relation = boundary (and =multipolygon) works fine for defining an area, and you can make holes to exclude at my “outparcels” or villages which are not part of the official protected area.



the difference is that in case a thing which is either outer or inner member of a boundary or multipolygon changes, you will not know whether the MP has to change or you have to duplicate the old border for the multipolygon because it should remain. In a site relation it is clearer that the modification of a member should likely have immediate effect on the relation as well.
 


Mappers don’t need to add things to relations when the geometry is enough to show that node A isi side of area B.


different semantics. IMHO there are situations where explicit relation is preferable over implicit spatial relation.
This said, I am hardly ever using site relations ;-)

Cheers,
Martin

_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
Reply | Threaded
Open this post in threaded view
|

Re: Tagging of State Parks in the US

Mateusz Konieczny-3



Jul 30, 2019, 9:54 AM by [hidden email]:


Am Di., 30. Juli 2019 um 00:51 Uhr schrieb Joseph Eisenberg <[hidden email]>:
-1 to a site relation for an area with a defined outer boundary.


Relation = boundary (and =multipolygon) works fine for defining an area, and you can make holes to exclude at my “outparcels” or villages which are not part of the official protected area.



the difference is that in case a thing which is either outer or inner member of a boundary or multipolygon changes, you will not know whether the MP has to change or you have to duplicate the old border for the multipolygon because it should remain. In a site relation it is clearer that the modification of a member should likely have immediate effect on the relation as well.
Can you give more specific example
of edit where site relation is better?

Things that are areas should be tagged
as areas. So multipolygon seems
preferable to me.


_______________________________________________
Tagging mailing list
[hidden email]
https://lists.openstreetmap.org/listinfo/tagging
123