Re: [Talk-us] Unintentional improvements in OSM data influencing / improving other databases

[Bradley White]

>
> I echo this sentiment exactly as having taken place in California and in
> my experiences with OSM.  This is most certainly a longer-term endeavor
> (over several, even many years), but improvements in alignments between
> data components which have been entered into OSM from my County GIS,
> GreenInfo.org's publishing its "CPAD" (California Protected Area Database,
> published semi-annually, see our wiki) and other sources HAVE INDEED
> resulted in data improvements:  OSM influences CPAD, resulting in data
> improvements, CPAD influenced County GIS data, resulting in data
> improvements, later versions of these (County GIS and CPAD) data influenced
> OSM all over again, resulting in data improvements...and upward, upward and
> upward the spiral of more accurate, better-aligning data goes:  both
> private and public.  OSM gets the results, so do others.  Win-win.  Taking
> OSM out of the equation by asserting "these data don't belong in OSM" stops
> this improvement pipeline (wholly unintentional on my part, but certainly
> noticed) in its tracks.  (Yes, some data belong in OSM, some don't).


I'm in strong agreement here. OSM provides a unique platform to synthesize
multiple data sources in combination with field observation to produce
something potentially better than any of these single sources are on their
own. Trying to produce an accurate and detailed map of the entire US
strictly off of field observation and satellite imagery is simply
infesible, especially in remote, unpopulated areas. Many government and
agency data sources are in conflict with each other over the same
information; OSM can serve to provide "resolved" versions that are
confirmed with ground observation where required.

I agree that we shouldn't be importing parcel data wholesale, as-is. But,
if real-life accuracy is important, the fact that much of the information
we are trying to add in OSM (protected areas, land use, access
restrictions) is differentiated along parcel boundaries is simply
unavoidable to me. If this information is in the public domain and
generally corroborates what is on the ground, so long as the data is worked
through manually to confirm accuracy, I don't see the problem with using
parcel information as a piece of the "puzzle" in producing an accurate and
informative map.
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Trouble with getting Superior National Forest

[Bradley White]

>Protect area and National Park boundaries were supposed to be less difficult 
>to confirm and more valid.

The NF administrative boundaries are basically impossible to verify
on-the-ground if that's the standard we are setting to demonstrate
verifiability. Typically, the only indication are the large welcoming
signs placed adjacent to major highways running through NF land, and
even then these typically aren't placed exactly on these boundaries.
For example, the administrative boundary for the Humboldt-Toiyabe NF
includes half the city of Reno, but none of this urban land is
protected at all, and there is zero on-the-ground indication of this
boundary.

> But if what we are going to start mapping in the USA is simply the federal 
> ownership of land, that's just pure cadastre data. We might as well try to 
> map all the private land parcels and keep that information accurate - but 
> both tasks are too difficult, and the data is better provided by local 
> governments directly.

I think this is a bit of a slippery slope argument. At least in
California, the NF land ownership boundaries are public record with no
copyright (https://wiki.openstreetmap.org/wiki/Contributors#United_States).
The boundaries of the federally-owned parcels *are* the protected
areas - you can't accurately map them without the parcel data here.
Using this data doesn't mean we have to start importing county parcel
data carte blanche.

If we shouldn't use land ownership because this relies on parcel data,
and the reason we don't use parcel data is because it is subject to
change and generally unverifiable on-the-ground, then we really
shouldn't be using NF administrative boundaries either since they are
likewise imported from easily accessed government data sources,
subject to change, and unverifiable on-the-ground.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Trouble with getting Superior National Forest

[Bradley White]

>
>  If you drive into a checkerboard
> area of private/public land, there are no Forest Service signs at the
> limits of private land.
>

In my neck of the woods, USFS owned land is signed fairly frequently with
small yellow property markers at the boundaries.

Privately owned land within a NF declared boundary is not under any
protection by the USFS, therefore tagging the administrative boundary as
'protected_area' will lead to inaccuracies. The land areas that are
actually protected from development/have active resource management are
only the lands which the federal government owns within these
administrative boundaries.

I think using the administrative boundaries is a good & practical first
approximation, but the goal should eventually to be to change over to the
actual land owned by the Fed and operated for conservation by the USFS.

>
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] National Forest boundaries

[Bradley White]

> We were doing great there, then I think my (admonishment?  might be too 
> strong) way of expressing "owned and operated by the USFS" is technically, 
> accurately stated as "owned by the People, managed / operated specifically by 
> the USFS."  If you can agree with me there, I think we can get even closer.

In most county assessor records, the name on the "title" of USFS owned
land is "United States of America", "United States Forest Service", or
some variant. The federal government owns the land, and manages the
land resource as well as US citizens' legal right to access the land
(barring conservation necessities that limit access to certain users
or any public at all).

> A USFS NF is a "virtual" multipolygon (not one in OSM, we can get to that 
> later) of three kinds of things:
>
> 1) An "outer" (but not the biggest one) which is "the enclosing land which 
> USFS manages, except for inholdings, below,"
> 2) Zero to many "inner" polygons, representing inholdings (and with the usual 
> "hole" semantic of exclusion from 1), above and
> 3) An even LARGER and ENCLOSING of 1) "outer" which Congress declares is the 
> geographic extent to which USFS may or might "have influence to someday 
> manage."

Sort of. Administratively, the USFS operates 9 regions containing 154
"national forests", with each forest being subdivided by a number of
ranger districts. The federal government also owns large swaths of
land across the country. These parcels are then managed by whichever
national forest (and ranger district) they happen to be located in.
There isn't necessarily an "outer" way enclosing the land that the
USFS manages, there is just a sum of US-owned parcels that fall within
a certain NF boundary that represents the actual land managed by the
USFS. In OSM practice, this is often a very complicated multipolygon
with multiple 'outer' members, which is usually required in order to
avoid self-intersecting rings.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] National Forest boundaries

[Bradley White]

> However, I'm not exactly sure how the outer polygons found in NFs differ from 
> either the "Congressional" boundary or the one Bradley says he would tag 
> "boundary=administrative" (and I don't think we should tag it that, 
> especially while excluding a specific value for admin_level), but I'm willing 
> to listen to more discussion about what this "different from Congressional" 
> boundary is and how the two differ.  Apologies if that isn't clear, I'm doing 
> my best, but I remain unclear on some concepts here.

NF congressionally designated boundary, minus private inholdings (more
specifically, non-USFS-owned land), gives you the boundary of land
that is actually managed and protected by the USFS. This boundary
should be tagged with 'protect_class=6'. USFS owned land is always a
subset of this congressional boundary (I suspect it is, in all cases
in the US, a proper subset). Subtracting these private inholdings is
generally going to change the shape of the 'outer' way such that it no
longer is the same as the "designated" boundary.

> My slight disagreement with Bradley is as above:  I don't think we should put 
> a "naked" (missing admin_level) boundary=administrative tag on these, it 
> simply feels wrong to do that.  (I READ the point that these are 
> "Congressionally designated" and that SEEMS administrative...but, hm...).

I wasn't clear in what I meant by suggesting 'boundary=administrative'
tagging here - I don't think we should tag "declared" boundaries
'boundary=administrative' with no 'admin_level'. This is simply the
closest widely-used tag that comes close to representing what this
"declared" boundary actually means. This is also why I suggest we
think about not including it at all in OSM; should we also start
adding boundaries for interstate USFS administrative regions (an
'admin_level', for lack of a better term, more general than a NF
boundary), as well as ranger districts within each national forest?

The real, on-the-ground objects of importance here are the plots of
land that are actually owned and operated by the USFS, not an
administrative boundary that declares where each national forest *may*
legally be authorized to own and manage land, and that is not
surveyable on the ground.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] National Forest boundaries

[Bradley White]

> Somewhat related, in the cases where an official FS road or trail crosses 
> private property, does the FS have an easement, or is it kind of an informal 
> arrangement?

Best way to know for sure is ground survey, but generally USFS system
roads & trails (also available for viewing using the USFS data extract
tool) over private land are public easements. If a section of the
system road/trail 'disappears' over a piece of land, it might not be
open to the public. An on-the-ground survey is usually required in
those cases.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] National Forest boundaries

[Bradley White]

> While it certainly may exist, I'm not aware of a disparity between the 
> "congressionally declared boundary" and any other boundary of a NF, including 
> "physical land that the NF actually owns and manages."  How would anyone know 
> where this latter boundary is?

The declared boundaries are administrative boundaries that limit the
extent in which each NF *may* manage land, but only land owned by the
USFS within these boundaries is actually protected at
'protect_class=6' criteria. Both of these boundaries are available for
download using the USFS Data Extract tool, and specifically in
California, the surface ownership boundary of each national forest is
included in the CPAD database. They can also usually be verified on
the ground by yellow NF property markers, as stated previously. In
fact, it is the congressionally declared boundary that is impossible
to verify on the ground.

Having lived in multiple places within a "declared" NF boundary, the
NF affords no protection on the land I have lived on. There might be
some extra hoops to jump through when pulling permits, but that
certainly isn't enough to include it within a 'protect_class=6'
boundary.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] National Forest boundaries

[Bradley White]

> A relation for all would be ok too, as long as the private inholdings are
> not removed from the NF (which I think has been done in some cases).

I've argued for this in the past on this mailing list, but have since
come around to disagreeing with this position over tagging semantics.
Most NF boundaries are now tagged with 'boundary=protected_area', in
which case the boundary should represent physical land that the NF
actually owns and manages, and not the congressionally-declared
boundary. In my area, half of the city of Reno and nearly all of
Truckee fall within an congress-declared/administrative NF boundary -
these areas are certainly not protected.

IMO, a tagging scheme that better represents the meaning of these two
boundaries would be:
1. 'boundary=protected_area' around fee simple NF land ownership,
since this describes the actual protected areas of land
2. 'boundary=administrative' (with a not-yet-existing 'admin_level')
around declared NF boundaries, since this is an administrative
boundary for the NF and doesn't necessarily show what land is actually
managed by the NF.

We should even consider not including congressionally-declared
boundaries, since they aren't even theoretically verifiable on the
ground, and really don't necessarily indicate any kind of protection
of the land within the boundary. Fee simple ownership is at least
usually ground-verifiable with small yellow "NF boundary" placards.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] United States Bicycle Route System ballots pending AASHTO approval

[Bradley White]

Completed Placerville to Folsom - couple questions.

Is the suggested segment along Tong Road accessible to the public?
It's a recommended "neighborhood connector" according to the Western
El Dorado County Bike Map and appears to see decent traffic according
to Strava heatmap, but the parcel map for El Dorado County doesn't
show it as a right-of-way, and the old (probably out of date!)
California Cross State Bicycle Route Study says that the roadway has
been gated off and was under investigation as to whether it was open
to the public or not. Signage at the end of Old Bass Lake road seems
to suggest the road is private (street view with
that-which-shall-not-be-named), but I have often seen these signs up
along a road where the land around the road is private but the road
itself is a public ROW.

Also, there is no physical crossing to turn EB from Silva Valley Pkwy
onto Tong Rd - one would be required to either dismount in the middle
of the road and walk, or bunny hop up over the median crossing four
lanes of traffic. There are also no crosswalks to circumnavigate the
barrier anywhere around. I have added a crossing here as a footway
with crossing_ref=none but this seems dubious to me and difficult to
tag appropriately.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] United States Bicycle Route System ballots pending AASHTO approval

[Bradley White]

>If you are in California (or even if not!) and want to enter USBR 50, helping 
>to build Earth's largest official cycling route network, check out our wiki, 
>follow the links to the turn-by-turn and map data and have fun!

Just finished adding the route from SLT to Placerville, plan to
continue west as I have time.

Aside: Curiously, the crossing proposed across US 50 at Sierra Blanca
(line 21 - confusing to follow since you do not "continue on" but
instead turn left and cross 50 to get to Sierra Blanca, and Ponderado
Rd. is not involved in the route at all per El Dorado County road map)
appears to be illegal, unless this somehow counts as an implicit
crossing under CA law. Not to mention having to cross 6 lanes of 65
mph+ traffic! I would prefer to stay on Carson Rd. until it reconnects
with El Dorado Trail myself.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Alaska Highway AK-2 tagging

[Bradley White]

> Long term, it would be nice to separate these notions and have some
> highway:importance key for that, and leave the road type notion that
> separates primary/trunk/motorway alone (or move it to some other tag,
> and get rid of highway=trunk and highway=motorway).

Ideally, this is what 'highway' is already supposed to represent
(https://wiki.openstreetmap.org/wiki/Proposed_features/Highway_key_voting_importance).
The debate fundamentally is over whether roads built to "expressway"
standards are de facto important enough to warrant being their own
class of importance, in the same way that freeways are in the US
(given that the US has a national interstate system). I think there
are convincing cases for both tagging schemes, but that ultimately we
need to make a decision as a community, since using both is ambiguous
and confusing.

> I guess you are suggesting to add highway=expressway to have expressway
> mean "sort of motorway but not quite" and change trunk to be "very
> important".  I am afraid that with so much established tagging the only
> reasonable approach to orthogonalization is to adopt two new tags for
> the things in question and deprecate the old way, allowing for a long
> and messy transition.

If 'trunk' became defined as "most important roads", I would suggest
adding a tag like 'expressway=yes/traditional/super_two/...', default
'no', to indicate the built design of the road. There has already been
such a proposal for a while, though currently abandoned:
https://wiki.openstreetmap.org/wiki/Proposed_features/Expressway_indication

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Alaska Highway AK-2 tagging

[Bradley White]

On Mon, Dec 16, 2019 at 4:53 PM  wrote:

> On the West Coast, several important State highways are tagged as trunks
> even though they are not full expressways, because they are the main road
> for a large region. For example, see US 199, US 101, CA 99 and CA 299 on
> this map of far Northern California:

FWIW - I am the one who bumped each of these roads listed up to trunk
a year or two ago, and I have recently bumped them back down to
primary (what they were for years before IIRC) to remain consistent
with guidelines posted here:
https://wiki.openstreetmap.org/wiki/United_States_roads_tagging and
here: https://wiki.openstreetmap.org/wiki/United_States/Road_classification

The lack of consistent highway tagging in the US is one of the biggest
sources of frustration with this project as a whole to me. IMO, the US
community needs to make a decision to *either*:

1. Use 'trunk' to mean "major cross-country highway" and orthogonalize
expressway constructions with its own 'expressway=*' tag, bump
'primary' to "minor cross-country highway/major regional highway",
'secondary' to "minor regional highway/major local road", etc...

Or:

2. Use 'trunk' to mean strictly "partially grade-separated limited
access divided highway" (with explicit instruction to not tag singular
or isolated interchanges as 'motorway')

The mixture of the two schemes that leans towards one or the other
depending on what part of the country you're in is inconsistent and
confusing to me, and judging by how many times we've gone in circles
about this on us-talk, others as well.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] National Forests and Private Ownership

[Bradley White]

> I downloaded a quad (geotiff) for part of the area in question and pulled it 
> into QGIS.  It generally agrees with the county land ownership information, 
> with the exception that some state lands are shown on the quad as owned by 
> the Federal Government.  Perhaps this is an error in one of the datasets.

As far as I understand (and tag), the owned lands should be tagged as
landuse (generally, landuse=forest, access=yes, operator=xxx National
Forest), but the administrative boundary is something different from
that. From what I see in parts of the SE (looking specifically at
Chattahoochee-Oconee, Nantahala, etc area), the administrative
boundary multipolygon is around *only* USFS owned lands. I would
consider this tagging style incorrect, and largely the exception
relative to the rest of the US.

>The Fee Owned is a a subset of the Congressionally Mandated boundaries as 
>someone else explained. My unofficial suggestion is if you want to model 
>recreation, it would be better to show the Fee Owned boundaries so people 
>don't end up on private lands. The US Topo uses proclaimed at this time.

Proclaimed boundaries are the administrative boundaries, and should be
tagged with "boundary" tags as they are in most of the US. Actual
owned land (fee owned) is a matter of landuse and should be tagged
using landuse tags. Unfortunately, this doesn't necessarily show
distinctly on the main OSM slippy map. However, this is a generic map
that is not *necessarily* designed to be useful for any one specific
thing, and trying to show different ownership of land can get messy
very fast. The data in the OSM database should reflect the distinction
between designated administrative boundaries and actual managed forest
land, regardless of whether this shows nicely on the slippy map.

If the goal is to see clearly what land is actually owned by the USFS,
then it is likely better to either use the USFS topo maps, or develop
your own map style that shows the difference!

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] National Forests and Private Ownership

[Bradley White]

> One point is that the NFS may have made arrangements with the landowner such 
> that some access by the public is permitted.  I say this because an official 
> USFS trail (Crosier Mountain Trail)[1] crosses private land and there are no 
> signs saying "No Trespassing"

The way may be, but usually the land itself is not. The land is still
tagged access=private, and the trail is tagged either access=yes if it
is a legal public easement over private land, or access=permissive if
there is an agreement with the landowner to allow public to access the
trail as long as they stay on trail (but there is no legal right of
way otherwise)

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] National Forests and Private Ownership

[Bradley White]

Sorry - not too familiar with imgur! Does this work?
https://i.imgur.com/4OC23x3.png

On Tue, Oct 15, 2019 at 1:24 PM Mike Thompson  wrote:
>
>
>
> On Tue, Oct 15, 2019 at 2:21 PM Bradley White  
> wrote:
>>
>> A visual example since I don't feel like what I'm saying is being
>> understood: https://imgur.com/a/0ELKyxH
>
> The link takes me to a page that is asking me to sign in.
>>
>>

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] National Forests and Private Ownership

[Bradley White]

Yes I understand that, that is what the landuse tag is for. Private
land should tagged as private. Public land should be tagged as public.
The 'access' tag is probably preferable for this, and it's what I use.
My point is that none of this involves the NF boundary, and to please
leave it alone because it's a pain to fix problems with it.

On Tue, Oct 15, 2019 at 1:22 PM Mike Thompson  wrote:
>
>
>
>>
>> Please do not add holes in the boundary unless they are officially
>> designated! Otherwise there is no point to keeping these
>> administrative boundaries in OSM.
>
> Ok, but we still need to know where those private inholdings are, because 
> Forest regulations will not apply.  For example, unless posted otherwise, I 
> can go anywhere on National Forest government owned lands, and I can camp 
> anywhere as long as I am not within a certain distance of a road or stream.  
> I can't do those things on private land. So access=private, ownership=private?
>

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] National Forests and Private Ownership

[Bradley White]

A visual example since I don't feel like what I'm saying is being
understood: https://imgur.com/a/0ELKyxH

This key works for anywhere on this
(https://data.fs.usda.gov/geodata/rastergateway/states-regions/states.php)
slippy map - take a look at the national forests near you and you will
find plenty of private land that is still within the NF boundary.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] National Forests and Private Ownership

[Bradley White]

Every National Forest has an administrative boundary - they can be
downloaded here:
https://data.fs.usda.gov/geodata/webapps/EDW_DataExtract/. Accept the
disclaimer, click the button with the scissors in the top left corner,
choose the national forest you want, select 'Administrative Forest
Boundaries' (preselected), choose your file format, and open in your
favorite GIS program. This boundary is what is in OSM, or at least
what should be. These boundaries can also be viewed using USFS Topo
maps (https://data.fs.usda.gov/geodata/rastergateway/states-regions/states.php)

You will see that sometimes private land punches a hole in these
boundaries, and if so it should be in OSM as such. But you will also
see that sometimes (often times in the west coast), private land
*doesn't* punch a hole in the boundary, and thus there *shouldn't* be
a hole in the boundary in OSM despite being a private in-holding. This
is what I mean by these conflating landuse and jurisdiction. Private
land inside NF boundaries does not automatically mean there's a hole
in NF boundary.

Please do not add holes in the boundary unless they are officially
designated! Otherwise there is no point to keeping these
administrative boundaries in OSM.

On Tue, Oct 15, 2019 at 12:45 PM Kevin  wrote:
>
> Bradley,
> I'm not sure that this is typically how federal lands are conceptualized, at 
> least on the east coast.  It is usually as Mike suggests a 1:1 correspondence 
> with the actual Fee Simple boundary and federal management.  A lot of times 
> when maps are drawn or gis data is developed scale is a consideration and 
> just conveying where a National Forest is is more important than showing a 
> patchwork of in-holdings (which by the way are constantly changing with land 
> swaps and selling or buying parcels). This may be where the idea of an 
> administrative boundary or area comes from? In any case a really excellent 
> source for all protected lands is the USGS PAD-US dataset.  
> https://www.usgs.gov/core-science-systems/science-analytics-and-synthesis/gap/science/protected-areas
> Disclaimer: I am the Georgia data steward.
>
> So Mike,
> I would say if you have the information and data that there's a private 
> in-holding, I would exclude it from the National Forest (or whatever) polygon 
> and maybe map the landcover (forest, etc) if you are so inclined.
>
> Kevin
>
> On Tue, Oct 15, 2019 at 3:12 PM Bradley White  
> wrote:
>>
>> No, this is incorrect. USFS administrative boundaries and USFS managed
>> land are not the same thing, though the latter is always inside the
>> former. The boundaries currently in OSM are administrative boundaries,
>> and are tagged correctly as such. It is perfectly fine to have private
>> land within a USFS administrative boundary, in the same way it would
>> be okay to have private land within any other government-defined
>> jurisdictional boundary.
>>
>> > The consensus of those who replied seem to be to exclude these privately 
>> > held lands from the National Forest boundaries.  Is that correct? Does 
>> > anyone object to that approach?  If not, I will proceed in that manner as 
>> > well.
>> >
>> > Mike
>> >
>>
>> ___
>> Talk-us mailing list
>> Talk-us@openstreetmap.org
>> https://lists.openstreetmap.org/listinfo/talk-us

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] National Forests and Private Ownership

[Bradley White]

> Ok, so how to tag the parts that are within the administrative boundaries but 
> which are not owned by the US Government? Or, how to tag the parts that are 
> both within the boundary and owned by the US Government?

It depends on what is actually on the ground. It appears you and
others are conflating jurisdictional boundaries with
landuse/ownership. While NF-owned land must be within a NF boundary,
that is the end of any relationship between NF boundaries and
on-the-ground landuse. The "National Forest property behind this sign"
demarcates landuse, not jurisdiction. For example, in theory, there
could exist a single parcel of private property, that is also
partially within a designated wilderness, that also spans across two
different national forest boundaries. There's no casual relationship
between these concepts, in the sense that "this land is private,
therefore it is a 'hole' in the NF boundary".

What is actually on the ground should be tagged using landuse.
Private forest cabin within NF? landuse=residential, access=private.
Tree-covered land owned by USFS? landuse=forest, access=yes,
operator=Tahoe National Forest.
Private timber harvesting land? landuse=forest,
access=private/permissive, operator=whoever.
Notice that none of these involve changes to anything 'boundary',
because they're distinct and (mostly) orthogonal concepts.

The NF boundaries, for the most part, are correct in OSM as they are
and should not be touched unless incorrect per USFS GIS data which is
the reference for them. It's difficult to notice when they've been
incorrectly changed, and it's even more difficult to fix them once
they have been messed up. Someone has made up a lot of work in
California a few years ago by making wilderness boundaries share ways
with NF boundaries (mutually excluding NF jurisdiction from wilderness
area), when in fact wilderness areas *overlap* NF jurisdictional
boundaries and do _not_ exclude them (ie, wilderness areas are often
managed by multiple National Forests, and are not their own
separately-managed entity). I have fixed a couple near Lake Tahoe, but
it is enormously time consuming work that requires some experience
with GIS tools as well as JOSM, which is very frustrating considering
they were correct in the first place.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] National Forests and Private Ownership

[Bradley White]

No, this is incorrect. USFS administrative boundaries and USFS managed
land are not the same thing, though the latter is always inside the
former. The boundaries currently in OSM are administrative boundaries,
and are tagged correctly as such. It is perfectly fine to have private
land within a USFS administrative boundary, in the same way it would
be okay to have private land within any other government-defined
jurisdictional boundary.

> The consensus of those who replied seem to be to exclude these privately held 
> lands from the National Forest boundaries.  Is that correct? Does anyone 
> object to that approach?  If not, I will proceed in that manner as well.
>
> Mike
>

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] National Forests and Private Ownership

[Bradley White]

> Not all of the land within US National Forests is owned by the US
> Government, there are private "inholdings" [1].
>
> The boundaries between government land and private land are often marked by
> signs, e.g.[2]  The above photo is geotagged, and if you drag it into JOSM
> you can see that it is quite far from the overall National Forest boundary
> as currently depicted in OSM[3].

Land actually owned and operated by the USFS is always a subset of the
jurisdictional boundary of a given NF. Near where I live, half of the
entire city of Reno is within the Humboldt-Toiyabe boundary, the
entire city of South Lake Tahoe within LTBMU, town of Truckee entirely
within Tahoe NF, etc. The jurisdictional boundaries are more or less
unhelpful in determining whether land is managed by the USFS or not.
I'm assuming this must not be the case in other parts of the country,
where the vast majority of the land within a boundary can assumed to
be owned by the USFS?

Aside from surveying boundary markers (which are inconsistently placed
and would be a logistically impossible task), the only other ways to
know what land is actually owned by the USFS is to check county parcel
data, or use the 'Surface Ownership' gdb/shp available using the USFS
Data Extract tool. In CA, we are very lucky to have the CPAD database,
which compiles the majority of public/semi-public lands into one
database, updated yearly, and free to use (see Contributors page in
OSM wiki). Where these lands also have tree cover, I tag them
'landuse=forest' and 'access=yes'. Any private "inholding" gets tagged
for what it is.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Historic 66 as highway=trunk in OK

[Bradley White]

> Also language introduced by NE2 when he changed the wiki to justify his own 
> national mass edit on the US highways.

If all this language was added unilaterally by NE2, can we find the
specific wiki edits that they made and roll them back? I'm on the same
page with Steve that describing how tagging currently *is* used and
debating over how tagging might be better used, should be kept
separate.

> Feels like conflating expressways and primaries.

Certainly, relative to current tag usage, which is why I'm debating
current tag usage. If 'highway' is supposed to, fundamentally, denote
importance (Proposed_features/Highway_key_voting_importance), it seems
to me that there's a big difference in importance between a highway
that connects population centers of ca. 10,000+ (current primary usage
as described in United_States_roads_tagging), and a highway connecting
population centers of ca. 100,000+ (Vegas to Boise, Reno to Redding,
etc), regardless of the physical condition of the road. Conversely, it
seems strange to use such a tag as high in the hierarchy as 'trunk'
for a fairly minor state highway that is otherwise 'secondary', simply
because the highway turns divided. I can sympathize with many mappers,
especially newer ones, not understanding what to do with 'trunk'. It
is probable that this 'strangeness' is due to rendering choices in
osm-carto, but it's been made clear that country-specific rendering is
logistically near-impossible and won't be happening. I've thought
about putting together a US-specific style that uses highway ref as a
factor in rendering, so that highways that are properly tagged per OSM
standards show up as one might "expect" them to on a US map. But I
certainly can't afford to host a tile server right now.

I'd be much more on-board with the 'trunk' = 'expressway' thing (been
halfway there for a bit) if the following happened:
- Removal of "important highway where no motorway exists" or
equivalent verbiage from wiki tagging guidelines. If this was added
unilaterally by one editor, it should be removed regardless.
- Rewriting 'trunk' section of US road tagging guidelines, with a
section on understanding the concept of access control, and including
multiple photographed examples of different kinds of expressways with
descriptions. I'd be happy to help contribute to this.
- Systematic review of 'primary' use in the US - if this is going to
mean 'nationally important road', there shouldn't be things like
nearly every state highway being tagged 'primary', or downtown areas
filled to the brim with 'primary' (Houston and LA as particularly
egregious examples), or 'primary' roads that almost exactly parallel
an interstate (the interstate is the primary road!). I have been
working on this in California for a while, but it usually quite
time-consuming - many roads have been inappropriately bumped up so
they cut through the TIGER mess, when in reality it's the TIGER mess
that needs to be cleaned up first.


> Feels like conflating expressways and primaries.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Historic 66 as highway=trunk in OK

[Bradley White]

> For example, US Hwy 101 is the main route connecting the cities (e.g.
> Eureka) and towns along the coast of northern California. Right now
> only some segments are tagged as highway=trunk. I would like to
> upgrade all of it to highway=trunk, up to Hwy 199, where most traffic
> leaves 101 and heads to I-5, at Crescent City.

I did this a year or two ago, then changed it back following the
previous time this discussion came up last year. Someone else has
recently changed it back to trunk in its entirety as you describe (as
well as US 395, CA 70); I explained in a changeset comment that the
"major intercity highway where no motorway exists" definition (per
Highway:International_equivalence) is contentious and not commonly
used, but that I have no plans on reverting their changes.

Caltrans doesn't appear to have "divided" as a requirement for an
expressway build, or even necessarily a freeway (See:(California)
State Highway Map 2005; David Rumsey Map Collection) - these terms are
used to describe the level of access control on a given highway. US
101 through Redwood Ntl Park is signed with "Freeway Entrance" and is
fully access controlled, but is an undivided 4-lane road. Many 2-lane,
undivided roads are considered expressways in California, for example:

- Vasco Road connecting Antioch & Livermore
- Portions of CA 4 west of Angels Camp
- CA 108 east of Sonora (fully access controlled 2-lane road)

Once you know what to look for - reduced access to adjacent
properties, smoothed road geometry (esp. when bypassing old highways),
hard shoulders, usually 65 mph - they aren't too hard to differentiate
from conventional 2-lane highways with no access control. Where these
are obvious I generally tag them as trunk roads as opposed to primary.
Specifically in the case of CA 108, I reject that a fully access
controlled two-lane road is anything less than a trunk, if we have
decided to use 'trunk' to mean 'expressway'. California doesn't use
AASHTO definitions so I won't either.

Reno, NV has a couple urban arteries that straddle the divide between
trunk and primary (specifically: McCarran Blvd/NV 659, Pyramid Hwy/NV
445 north of McCarran, Veterans Pkwy, foothills portion of Mt. Rose
Hwy/NV 431). These roads carry traffic at speeds higher than other
nearby arteries (45-55 mph as opposed to 40 mph). They are built to
the highest level of access control specified by Washoe RTC -
generally no direct access to properties, except for retail/commercial
areas (where access is quite frequent), or rural areas where no other
roads provide access to properties. They range from undivided w/
center turn lane to divided with concrete jersey barriers & headlight
blinders (similar to a freeway). The majority of these roadways have
bike lanes, and many have sidewalks. They are quite similar to San
Jose's expressway system, except for a lack of grade-separated
interchanges. Are these primary, or trunk? I don't really know. They
currently sit at an awkward mix of trunk and primary depending on how
definitively myself and others think they are "expressways" or not.

I don't deny that "divided highway with partial control of access" is
a rigorous definition, with which it is certainly possible to tag
unambiguously with. I just question whether it is a good choice in the
US to use 'trunk' to mean 'expressway' in the same way that 'motorway'
means 'freeway', when the US has a formal freeway system, but lacks a
formal expressway system. Most other countries that also lack a formal
expressway system do not use the trunk/expressway definition (UK,
Canada, etc). In my area, sticking strictly to "divided highway with
partial control of access" means very few highways at all will see
'trunk' tagging. Certainly, this reflects what's on the ground here if
we use this definition - but why use a definition that either has to
be used ambiguously or seldom at all?

I support orthogonalizing expressways & trunk by using
'expressway=yes/no' for access control (maybe
access_control=full/partial/no?), 'highway=trunk' to mean non-freeway
road with national-level importance, and using 'oneway' to denote
whether a highway is divided or not. Then let rendering decide how to
draw the road from there. Want to see formal expressways drawn
separately? 'Expressway=yes' & 'oneway=yes'. Want a more general view
of the most important US highways? 'Highway=trunk'.

As it stands, I will continue to use 'trunk' on any section of highway
that is somewhere between a freeway and a conventional 2-lane highway
per US consensus. Hopefully one of these days consensus will shift.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Trunk versus motorway

[Bradley White]

> Can I get some voice of reason in
> https://www.openstreetmap.org/changeset/64919426?  There seems to be quite
> a few people (and one AARoads forum troll egging it on) that are trying to
> propel the idea that motorways have at-grade intersections, which is
> obviously incorrect.

I know I'm not going to change your mind, but I'd like to agree with
the other voices here that this scheme is overly pedantic without any
real justification for being so. No-one is saying that the motorway
has an at-grade intersection as you assert; the motorway simply
begins/ends *at* that intersection. Tagging freeway ending/beginnings
with this scheme is definitely not standard practice in the US, and I
don't see how changing this just to split hairs over freeway
definitions would benefit anyone.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] California is too big ;)

[Bradley White]

> 3. what would be a sensible way to split California - in 58 counties, or
> maybe just go with SoCal and NorCal for now?

I would suggest splitting into North & South along the northern edge
of the SLO/Kern/San Bernardino county lines as the first step; this
will at least split the LA and SF Bay areas into separate files, both
of which I assume account for a significant portion of CA's data.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Parks, again

[Bradley White]

> As you say "feel like Type 2" I think is where it fuzzies in my mind.  Parks 
> go to 3, 4, even 11 and beyond.  Parks have a wide range of "experiences" 
> besides 1 and 2.

So do roads. There are countless kinds of roads, with varying levels
of importance and physical features. Instead of using a catch-all
"highway=road" tag, and instead of tagging infinitesimal levels of
network importance (or any of the other countless possible metrics),
we develop a classification system that allocates all roads into a
small set of (semi)-easy-to-work-with-and-understand classes. Some
roads don't fit well into this system, true. It isn't always clean; it
can be ambiguous; it continues to be debated over, and that's fine.
But, for the most part, it has worked, certainly better than the
all-or-nothing alternatives would have.

I agree with previous posters that this is same case with parks. In
the same way that the fact that there is something different enough
about a freeway and a narrow county back-road to represent them
differently in the database, there is something different enough about
a park I would take a kid to play on the playground for an hour, and a
park that I can spend half the day mountain biking around in without
encountering more than a small handful of people, that I think they
should be differentiated between in our data. I don't think the title
given to a piece of land should necessarily have bearing on the data
representation, in the same way "Hampstead Heath" doesn't get
"natural=heath" just because it's in the name.

Currently, I use the tagging scheme detailed by Greg earlier. I am
certainly not opposed to using "leisure=park" along with a basic
classification tag, say "park=developed/undeveloped" or something, but
Greg's scheme has the benefit of using established tags with rendering
support that still more or less respect the definition and intent of
the tags. While "leisure=nature_reserve" has generally assumed some
kind of conservation status, I think the newish
"boundary=protected_area" tags do a much better job detailing land
conservation, and that "leisure=nature_reserve" is the perfect tag to
adopt for the type 1 parks which Greg talks about. These 'type 1'
parks are, after all, pieces of *nature* being *reserved* by a
government agency for *leisure* of the public.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Trunk

[Bradley White]

If we can determine importance (which is what the 'highway=' tag
fundamentally represents per the wiki) solely by what's on the ground,
why not just tag what's physically there, ditch the 'highway' tag
altogether, and let the renders handle it with their own algorithms?

>On Sun, Oct 15, 2017 at 12:19 AM, Paul Johnson  wrote:
>>
>> The US is pretty well known for overbuilding highways.  Are we trying to
>> document how things are on the ground or how things are actually
>> connected?  If we're going for the former, then yeah, only Bend Parkway and
>> a brief streak through Klamath Falls is a trunk part of US 97.  If we're
>> going for the latter, then go ahead with NE2's idea and smash almost
>> everything into trunk.
>>
>
>
>Keep hitting send too soon.  Personally, I find what's on the ground to be
>more useful than the connections.  Game theory and any routing engine can
>figure out the connections.  But knowing what's a stupid rural road with an
>overly generous speed limit and what's almost but not quite a freeway is
>more useful.  If I'm driving a big rig going from southwestern Canada or
>Alaska to somewhere in Nevada, I don't give two shakes what some toolbag
>things is the most prominent road.  I care more about what *actually is a
>big road*.  Calling a two leg segment of US 97 30km outside of East
>Butthump, Oregon a trunk is a great disservice when it's basically on par
>with County Road Number Who Even Cares tracing off to Outer
>Smalltownsville, other than the fact that it goes through.  Calling it a
>trunk when it's not is going to set an unreasonably high expectation for
>what is otherwise an overtravelled, glorified two digit National Forest
>route through the east Cascades frontier.  Primary is definitely ample for
>that road, even if you're going a more obscure minor haul route like Salem
>to Reno.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Trunk

[Bradley White]

The linked example is an OSM screenshot? So yes, especially if it is
strictly adhering to trunk==expressway, then they will be explicitly
marked. This is circular. USGS maps emphasize roads when they are
multi-lane highways that aren't freeways, not when they are
expressways. Not every multi-lane highway is an expressway, and not
every multi-lane highway is a trunk road.

On Sat, Oct 14, 2017 at 1:37 PM, Paul Johnson <ba...@ursamundi.org> wrote:
>
>
> On Sat, Oct 14, 2017 at 3:19 PM, Bradley White <theangrytom...@gmail.com>
> wrote:
>>
>> On Sat, Oct 14, 2017 at 12:53 PM, Nathan Mills <nat...@nwacg.net> wrote:
>> > Road maps in the US have long differentiated between freeway/expressway
>> > and
>> > has had both of those clearly different than US and state highways we'd
>> > be
>> > tagging as primary. Map users expect to see expressways shown
>> > differently.
>>
>> Could you show me an example of a US road atlas that explicitly
>> demarcates expressways? I have legitimately tried to find one but have
>> not been able to. Most US maps I've seen show freeways & toll roads
>> explicitly, but not expressways. Some maps might use a different
>> casing style to denote a divided highway, but the underlying color of
>> the line still represents the importance of the road. Which is the
>> point I'm trying to get at, that a highway being divided or not is
>> orthogonal to its importance.
>
>
> Just googling for it, I do find
> https://upload.wikimedia.org/wikipedia/commons/b/b3/Portland_map.png which
> has 99E where McLoughlin Boulevard is an expressway, and the Milwaukie
> Expressway, in green.  USGS's topo maps of Portland also show the Milwaukie
> Expressway as an expressway, though also shows Interstate 205 as an
> expressway instead of a freeway (which I think we'd all agree would be
> incorrect).  USGS also shows (albeit very outdated at this point) US 412 as
> being an expressway, different from a freeway, east of where the (then still
> unbuilt) Creek Turnpike now joins the Rogers Turnpike.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Trunk

[Bradley White]

On Sat, Oct 14, 2017 at 12:53 PM, Nathan Mills  wrote:
> Road maps in the US have long differentiated between freeway/expressway and
> has had both of those clearly different than US and state highways we'd be
> tagging as primary. Map users expect to see expressways shown differently.

Could you show me an example of a US road atlas that explicitly
demarcates expressways? I have legitimately tried to find one but have
not been able to. Most US maps I've seen show freeways & toll roads
explicitly, but not expressways. Some maps might use a different
casing style to denote a divided highway, but the underlying color of
the line still represents the importance of the road. Which is the
point I'm trying to get at, that a highway being divided or not is
orthogonal to its importance.

> It's less work on so many levels also. Creating a new tag requires
> significant work on the render side, but doesn't really gain us much over
> just using primary for roads that some think are important enough to be a
> trunk but are undivided. The wiki definition for primary is already "the
> most important non-motorway route between two cities" (essentially, and
> ignoring the variation in use between rural and urban areas)

This is incorrect. 'Tag:highway=primary' gives "A highway linking
large towns". 'Tag:highway=trunk' gives "Important roads that are not
motorways", and explicitly lists in the US tagging application that "a
major intercity highway" is a correct use of the trunk tag in the US.
'Highway:international_equivalence' as well as 'Key:highway' give the
same set of definitions. Where are you seeing that primary is "the
most important non-motorway route between two cities"?

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Trunk

[Bradley White]

I use Osmand frequently; the point of the cased-line style of the
trunk & motorway tags is, agreeing with Paul here, to show some degree
of access control. This is in-line with many paper road atlases,
especially older ones. My point was that third-party applications
choosing to use this style is their own pejorative, and we should not
be basing tagging definitions on how third-party apps use the data. In
regard to the trunk debate, I understand and fully respect Paul's
position, but I personally disagree. I'm hoping the debate here will
encourage the US OSM community in getting closer to an agreeable
definition for trunk.

On Sat, Oct 14, 2017 at 10:49 AM, Evin Fairchild <evindf...@gmail.com> wrote:
> To add onto what Bradley was saying about third-party applications, I just
> want to add that I've done some fact-checking about a claim that Paul made
> in a previous email about how Osmand renders trunks under the assumption
> that they are expressways (to be clear, by this I mean divided highways w/
> at-grade intersections). After some fact-checking, this claim receives a
> truth rating of completely FALSE.
>
> Anyway, I looked at how Osmand renders motorways versus trunk and I don't
> know how it is that you, Paul, can say that trunk is assumed to be like an
> expressway  in Osmand's render. That is simply not true. The motorway in
> Osmand, for those who are unfamiliar, is red with a thin blue outline around
> it, whereas trunk is just an orange-red line without any other color
> outlining it. This makes it look more like a single-carriageway road and
> less like an expressway like Paul falsely claims. All it looks like is a
> road that is of higher-importance than primary, and does NOT at all look
> like it could be an expressway. Usually, when maps show a divided highway w/
> at-grade intersections, it looks similar to a freeway, but a different
> color, whereas an undivided two-lane road typically looks nothing like an
> expressway or freeway. Thus, it is complete and utter lie to say that Osmand
> makes the assumption that trunk roads are expressways. I don't know how
> mkgmap shows trunk vs. motorway since I don't have a Garmin and thus cannot
> test it out, but I don't trust that Paul is telling the truth here either.
>
> It's important to make truthful claims here, Paul; from now on, I will have
> a VERY difficult time trusting anything you say. I know what I brought up
> was kind of a side point, but I think it's important to call out BS when I
> see it.
>
> -Evin (compdude)
>
>
> On Sat, Oct 14, 2017 at 10:23 AM, Bradley White <theangrytom...@gmail.com>
> wrote:
>>
>> > The concept of expressway and freeway are reasonably well known
>> > concepts;
>> > it makes a lot of sense to map trunk and motorway to those concepts.
>>
>> I agree with freeways but not with expressways. I have no data to back
>> this claim up, but I'm fairly convinced that, while the average
>> citizen could easily differentiate between "freeway" and "not
>> freeway", they would be hard pressed to do the same with an
>> expressway. Anecdotal, but even when I spent time in the Santa Clara
>> area which has a robust expressway system, I never heard a single
>> person say "and then get on the expressway...", or even the word
>> 'expressway' mentioned outside of it being the suffix of a road name.
>> You're right that it's not a terribly difficult concept to understand
>> and thus map, but I disagree that it's an important concept in
>> explaining the road hierarchy in the US, so much so that we can equate
>> an entire class of importance with them. We have a robust, clearly
>> signposted freeway network in the US. We do not have the same with
>> expressways. Roads tend to go in and out of "expressway" qualification
>> depending on context, traffic levels of connecting roads, and highway
>> budget & design policy. A road being built as an expressway is
>> suggestive of its importance at best, and certainly not indicative.
>>
>> Edmonton has many roads around the east and west of the downtown area
>> that are clearly built as expressways. However, they are only tagged
>> secondary because, fundamentally, you only really need to use them to
>> get around the immediate vicinity. Despite being very high quality
>> roads, they aren't all that important in the grand scheme. I can point
>> to many examples of urban roads that likely meet an expressway
>> definition in my current home city of Reno, including one under
>> construction. It would be absurd to me to tag them as being second in
>> importance only to motorways just because they are well-built roads,
>

Re: [Talk-us] Trunk

[Bradley White]

> The concept of expressway and freeway are reasonably well known concepts;
> it makes a lot of sense to map trunk and motorway to those concepts.

I agree with freeways but not with expressways. I have no data to back
this claim up, but I'm fairly convinced that, while the average
citizen could easily differentiate between "freeway" and "not
freeway", they would be hard pressed to do the same with an
expressway. Anecdotal, but even when I spent time in the Santa Clara
area which has a robust expressway system, I never heard a single
person say "and then get on the expressway...", or even the word
'expressway' mentioned outside of it being the suffix of a road name.
You're right that it's not a terribly difficult concept to understand
and thus map, but I disagree that it's an important concept in
explaining the road hierarchy in the US, so much so that we can equate
an entire class of importance with them. We have a robust, clearly
signposted freeway network in the US. We do not have the same with
expressways. Roads tend to go in and out of "expressway" qualification
depending on context, traffic levels of connecting roads, and highway
budget & design policy. A road being built as an expressway is
suggestive of its importance at best, and certainly not indicative.

Edmonton has many roads around the east and west of the downtown area
that are clearly built as expressways. However, they are only tagged
secondary because, fundamentally, you only really need to use them to
get around the immediate vicinity. Despite being very high quality
roads, they aren't all that important in the grand scheme. I can point
to many examples of urban roads that likely meet an expressway
definition in my current home city of Reno, including one under
construction. It would be absurd to me to tag them as being second in
importance only to motorways just because they are well-built roads,
because they're unimportant outside of getting around the relatively
small Reno-Sparks metropolitan area.

The "highway" key is about importance. The only category we have
full-stop made equivalent with a type of road design is "motorway".
From trunk on down, it is just different grades of importance. These
are how the definitions are listed on the 'Key:highway' page, which I
consider to be definitive. The fact that the words "trunk", "primary",
"secondary", ... are used is an artifact of the UK roots of OSM. Had
this project started in the US, the keys would probably be "freeway",
"principal_artery", "major_artery", "minor_artery", "major_collector",
... leaving UK users scratching their heads trying to figure out how
to adapt these definitions to their own network. In countries with
signposted expressway systems, it is meaningful in understanding the
road network to equate trunk with expressway, so they do that. I don't
think doing the same is meaningful in the U.S. given how much
variability and inconsistency there is with how and where expressways
are constructed.

> Even a lot of renderers make this same assumption:  mkgmap maps trunk to
> Garmin's concept of expressway and motorway to freeway.  Osmand, easily the
> most popular data consumer for OpenStreetMap, makes the same assumption (to
> the point that most of it's map painting styles, the only differentiation
> between trunk and motorway is a color pallette shift).  It really wouldn't
> hurt the US community to have a "come to Jesus" moment on this,
> particularly when using the MUTCD definitions for expressway and freeway as
> qualifiers for trunk and freeway, makes this relatively easy.  The
> corollary to "don't tag for the renderer" is "don't break the renderer".
> Highways without access control being excluded from trunk or motorway isn't
> an intrinsically bad assumption to make.  Especially if we come to
> agreement on that, we can start having a productive talk on how to make
> carto not suck for Americans without breaking it for everyone else.

I'm really not that concerned with how third-party applications decide
to paint their roads. It's up to them to work with the data we
provide, not the other way around. If it is important to Garmin or
other applications to translate expressways, this can usually be
deduced from other tags, or we can trivially add an "expressway=" tag.
I also disagree that the carto in the US is bad, other than our
insistence that two-lane are categorically not trunk leaving
meaningless splatters of orange around the map at low zoom.

Also, apologies ahead of time if I keep breaking the archive
hierarchy, I'm not totally familiar with how to drive a mailing list
and I have yet to find a guide online that explains how.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Trunk

[Bradley White]

> Message: 4
> Date: Fri, 13 Oct 2017 21:24:20 -0500
> From: Paul Johnson 
> To: OpenStreetMap talk-us list 
> Subject: Re: [Talk-us] Trunk
> Message-ID:
> 

Re: [Talk-us] Trunk

[Bradley White]

Lots of words ahead, you have been warned...

I disagree with trying to use the "highway=" tag to describe what
"kind" of road a given way is in the US, except for freeways. The
"highway" key is for importance, or, how prominently a road should
show on the map. We have other tags to describe completely the
physical attributes of a road. If there existed an "if-then" algorithm
to determine how prominently a road should show on the map only from
physical attributes, we wouldn't need this tag at all.

But we do. The U.S. needs this tag because form often does not follow
function. There are many plain-old two-lane roads that are important
enough for cross-country navigation that they should show at low zoom
along with the interstate system. There are also countless high-speed,
high-access-control (no driveways, limited intersections), multi-lane,
divided arteries that do little more than connect suburbs to more
important roads. These roads do NOT need to show at low zoom, despite
being a high-quality road.

Consider instead that we trade off "highway=motorway" with
"highway=1", "highway=trunk" with "highway=2", etc., which represents
only the importance of a road in the network. In the US, it is fair to
take freeways as an entire class to be the most important roads. A
freeway has strict & verifiable physical criteria (aside from a small
fringe set of exceptions), they are signposted and unambiguous, it is
a term well-understood in common vernacular that the average map
consumer would expect to see shown on a map, and they are nearly
always THE most important roads in the area. It is easy to say both
that "this is a freeway" and that "these are the most important kinds
of roads". In fact, this is the case in nearly all developed countries
on the planet. So, we instead define "highway=1" to just be
"highway=motorway" since it is nearly always the same thing.

In Europe, it would also be fair to use "highway=2" to simply
represent expressways as a class. Expressways (in the countries that
have expressway systems) are built to a verifiable design criteria,
are signposted and unambiguous (see:
https://en.wikipedia.org/wiki/Limited-access_road), are something the
average map consumer in the area is both familiar with and would
expect to see on a map, and are nearly always second in importance to
the motorway system. Again, it is easy to say both that "this is an
expressway" and "this is the second most important kind of roads". So,
in countries that have designated expressway systems, they define
"highway=2" to just be "highway=trunk" since it is nearly always the
same thing.

This situation is NOT the case for the majority of the US, and trying
to use this definition is what has been leading to unresolved
confusion about the purpose of the trunk tag. MUTCD gives a definition
of "divided highway with partial control of access". This is rather
vague, and as stated above, means countless unimportant suburban
arteries would now be considered the second most important roads in
the country. Many states haven't even adopted usage of this term at
all. I have seen planning documents of some counties that have
multiple grades of "expressways" depending on intersection distance,
speed limits, etc. Outside of bona-fide urban expressway systems like
the Santa Clara Expressway system, I think it's a nearly meaningless
term.

I have heard two kinds of attempts to describe what constitutes an
expressway and thus a trunk road in the US. The first attempt is the
"it's like a freeway but only kind of" definition. I don't like this
definition because if we're going to trade off an entire class of
importance with an entire class of road design, we should at least be
perfectly clear about what kinds of roads we are talking about, as we
are with freeways. The second attempt is to establish some kind of
verifiable physical criteria: divided, minimum 45 mph speed limit,
limited intersections, maybe has grade-separated interchanges. There
are also many problems with this approach. As discussed above, it
grades swaths of overbuilt roads as being more important that they
actually are. It makes roads that are crucial for navigation but don't
meet an arbitrary checklist difficult to pick out from the sea of
primary roads that the rural US currently exhibits. Furthermore, it
leads to the current situation of random splotches of deep-orange
lines visible at the same level as the interstate system scattered
across the US, which provides absolutely no meaning to the average
US-based map consumer. Being frank, I can't understand at all why
anyone considers the current state of trunk road tagging in rural
parts of the US desirable or useful or illuminating at all.

I propose defining trunk in the US to mean, formally, "The most
important non-motorway roads in the country". An extended definition
for the US follows below:

--
Trunk roads are the most important non-motorway roads in the country.
These roads connect major 

Re: [Talk-us] "System Continuity" in the Functional Classification network

[Bradley White]

> In this document is a concept called "System Continuity". In few words, a 
> roadway of a higher classification should not connect to a single roadway of 
> a lower classification, so the network remains interconnected.
> Do you know if this concept applies to OSM roads network also?

Motorway is purely a physical tag, so it does not necessarily follow
this principle. Primary, secondary, etc. should with few exceptions,
and usually do. Trunk, not so much. Most (probably) mappers in the US
use it to mean "expressway", and have arbitrarily varying cutoffs for
what constitutes as such, ranging from sensical (Santa Clara
expressway system, CA), to not so much (this U.S. highway has gone
from 5 lanes single-carriage to 4 lanes dual-carriage and thus has
become a trunk; this extremely important cross-country highway slows
down briefly through a small town and thus is no longer trunk).

Without getting too much into my opinions about this, the majority of
U.S. mappers use trunk to mean "almost a motorway but not quite", in
which case you will _not_ find continuity, while some U.S. mappers use
trunk to mean "most important roads that don't meet strict motorway
standards", in which case you will (or at least should) find
continuity.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

[Talk-us] The Republic of Molossia (and other micro-nations)

[Bradley White]

Something a little bit different:

The Republic of Molossia is a self-declared "micro-nation" located
near Dayton, NV, landlocked by the United States. The nation claims
full sovereignty from the United States; however, it is recognized by
neither the United States, nor any other country on Earth, as an
independent nation. You have probably heard about it before, since it
is one of the best-known examples of such a micro-nation in the US.

Within the past few months, this "nation" has popped into OSM,
complete with sloppily implemented "admin_level=2" and
"boundary=national" tags, view-able here:
http://www.openstreetmap.org/#map=18/39.32281/-119.53908. My
discussion point is whether this is a valid use of these tags. A
handful of quick searches about this topic didn't turn up anything for
me, so I'm assuming no precedent has been set yet. It is worth noting
that this is not the only micro-nation in the US.

I'm not inclined to think these tags are valid. Otherwise, there's
nothing stopping me from calling my backyard its own nation, slapping
together a wikipedia article, and entering it into OSM as a
full-fledged nation. However, since they are still geographically
based entities of interest to the public, I think they are worth
mapping

There is a proposal for disputed boundaries, but I don't think that's
valid either since there isn't really a dispute. The nation has gone
unacknowledged by the United States, and nothing has gone through the
legal process between the two nations (that I'm aware of) that could
constitute a "dispute". No other boundary tag is really applicable,
maybe a new "boundary=micronation" would work? De facto, US law still
applies in these "micronations", along with the law of whatever
jurisdictions the micronation belongs to, so I don't think an
admin_level tag is applicable.

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Combined parking/bike lanes

[Bradley White]

> Hi all. Has anyone worked out a good tagging scheme for combined
> bike/parking lanes? I'm not sure how common they are elsewhere but there
> are a number of such facilities in my city.
>
> For reference, you can see an example here:
> https://www.mapillary.com/app/?lat=43.06056159=-89.45121134=17=XByvWxyrk9quLK-noyoB5g=photo=mapbox_streets
> Notice the bike lane sign above the speed limit sign and the cars parked on
> the side. These are also accompanied with pavement markings indicating it
> is a bicycle facility. In effect it's like a regular bike lane next to a
> parking lane, but there's no stripe to separate the two.

These types of lanes are relatively common in parts of northern
California as well; since the physical space is still set aside for
both parking and cycling, and the only difference is the inner line of
paint (which is more a "stylistic" choice on part of the agency), I'm
not convinced this needs special tagging. The tags suggested earlier
are what I would use:

> parking:lane:(right/left) = parallel
> cycleway:(left/right) = lane

___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] Tagging National Forests

[Bradley White]

Just to add my two cents, I do not think that "landuse=forest" should be
tagged with national forest boundaries. That something is within a national
forest boundary does not guarantee that it is a managed forest, or even
that it has tree cover. A 'national forest' is more an administrative
boundary to me than anything - it designates an area with active federal
management and a stricter set of laws involving development, etc. Half of
Reno, NV where I reside is technically inside the Humboldt-Toiyabe National
Forest boundary, including the urban center. There is certainly nothing
that qualifies as a 'forest' here in the traditional sense. Even many
parcels just outside of urbanized areas of Reno that are both within the
national forest boundary and owned by the forest service have no tree cover
whatsoever, and couldn't possibly qualify for any definition of a forest
involving trees.

Personally I think the problem here is a poor definition of
'landuse=forest'. Does this mean land used for timber production? I see a
lot of on-the-ground verifiability issues with that sort of definition.
Should it imply a large, managed area of trees? As explained earlier, there
are many federally owned and managed 'national forest' areas with no tree
cover whatsoever. I would be partial to a definition of 'land owned
directly managed by a forestry service' - forestry land - but then mapping
something like that would require parcel-level imports since not every
piece of land owned by the forest service is clearly marked on the ground.

I personally only use 'natural=wood' anymore, since at the very least it is
easy to verify that trees exist. I don't care much for the 'original
growth' definition of 'natural=tree' either due to verifiability issues.
Much of the Lake Tahoe is second-growth forest, but without a forestry
degree I don't see the average mapper being able to tell where
second-growth starts and stops.
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us