Re: [Talk-us] guidelines regarding roads access

[Greg Morgan]

On Sep 24, 2017 8:33 AM, "Greg Troxel"  wrote:


Adam Franco  writes:

> One additional note is that at least in my area, the TIGER import
> incorrectly added access=private to many driveways and privately
maintained
> residential roads. Upon surveying these I've found that they are signed
> "Private" or "PVT" on the street-name sign to indicate
> private-maintenance/ownership (don't complain to the town about a lack of
> snow-plowing/grading), but do not in reality have an access restriction.


https://www.mapillary.com/app/?focus=photo=oyEdoFPUg3yayTv1deHZUQ=33.567826=-112.016051=17

I initially marked a street like this as access=private.  Then I found out
that this definition of private is more like Adam's experience. However, I
did not experience the issue with the TIGER import. I am not sure that an
HOA is a good thing.  I stay away from them for a number of reasons
including the leftovers of the subprime rate debacle.

http://www.azcentral.com/story/news/local/arizona-investigations/2017/09/14/phoenix-area-homeowners-associations-foreclosing-record-number-homeowners/595816001/


For a "private way" (legal term in my state for what I think you refer
to as "privately maintained residential road"), I agree that there
shouldn't be access=private.

For a driveway to someone's house, access=private seems right, in that
it's generally at least impolite to use that road other than as
visitor/delivery/etc.

...

But, that tends to lead to pink blobs in rendering, and I'm not sure
that's the right thing, as service roads having the status "you should
use these only when dealing with the adjacent entities" seems to be the
default/normal case.  We should adjust rendering, not access, to make
this pleasing.


The rendering has been changed to a gray dashed center line.

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

Re: [Talk-us] dubious church node

[Max Erickson]

>One more thing to know about GNIS: entries are never deleted.

One minor exception to this is if they determine that a given feature
has 2 IDs, one of the IDs will often be removed.


Max

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

Re: [Talk-us] dubious church node

[Martijn van Exel]

I sense a topic for the next Many Mappy Minutes or a BOF at State of the Map US 
—> cleaning up old imports.
I sure appreciate all the knowledge shared here! So much I didn’t know about 
GNIS data.
Martijn

> On Sep 30, 2017, at 11:54 AM, Wolfgang Zenker  
> wrote:
> 
> * Carl Anderson > 
> [170930 17:21]:
>> ​A little history on GNIS data, and the Board of Geographic Names.
> 
>> The US Board of Geographic Names manages names for places and features
>> shown on US govt maps.  They have been using a database to manage the names
>> across maps and map scales. That database is the GNIS.
> 
>> The ​original GNIS data was populated from all text labels shown on USGS
>> maps.  The most common source was 1:24,000 scale topo quarter quads.  Text
>> from 1:100,000, 1:250,000 and 1:1,000,000 scale maps and larger were
>> included.
> 
>> The stated map accuracy of these scales  (
>> https://nationalmap.gov/standards/nmas.html ) is approximately
> 
>> 1:24:00040 feet
>> 1:250,000 416 feet
>> 1:500,000 833 feet
>> 1:1,000,000   1666 feet
> 
>> The GNIS dataset includes the most precise location for text, when the text
>> appears on maps of different scales.
> 
> You can look at the full database entry for an individual GNIS feature
> if you search for the GNIS Feature ID at geonames.usgs.gov/apex/f?p=gnispq 
> 
> 
> This will give you the source of the database entry, possibly a list of
> alternate names, sometimes a note like "location approximate", and
> sometimes the history of the decision process if more than one name
> had been proposed for the feature. Also documentation of official
> name changes.
> 
> One more thing to know about GNIS: entries are never deleted. If a
> feature no longer exists, the name gets "(historical)" appended to
> it. This may have happened after the feature was imported to OSM,
> so it may not show in the OSM database.
> 
> Unfortunately the GNIS database is no longer fully maintained due to
> budget constraints, so you can't be sure if features still exist even
> if they are not flagged as "(historical)".
> 
> As to mapping in OSM: I usually remove any "(historical)" feature.
> For the others, I improve the location if possible, and if the feature
> can be represented as an area, I draw that area/polygon.
> Instead of deleting the original POI, I now reuse that node as part
> of the outline of the feature and only move the tags to the area, so
> someone looking at the object details can notice that one of the nodes
> is a lot older than the others and still find the osm history of the
> feature on that node.
> 
> Wolfgang
> ( lyx @ OSM )
> 
> ___
> 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] dubious church node

[Wolfgang Zenker]

* Carl Anderson  [170930 17:21]:
> ​A little history on GNIS data, and the Board of Geographic Names.

> The US Board of Geographic Names manages names for places and features
> shown on US govt maps.  They have been using a database to manage the names
> across maps and map scales. That database is the GNIS.

> The ​original GNIS data was populated from all text labels shown on USGS
> maps.  The most common source was 1:24,000 scale topo quarter quads.  Text
> from 1:100,000, 1:250,000 and 1:1,000,000 scale maps and larger were
> included.

> The stated map accuracy of these scales  (
> https://nationalmap.gov/standards/nmas.html ) is approximately

> 1:24:00040 feet
> 1:250,000 416 feet
> 1:500,000 833 feet
> 1:1,000,000   1666 feet

> The GNIS dataset includes the most precise location for text, when the text
> appears on maps of different scales.

You can look at the full database entry for an individual GNIS feature
if you search for the GNIS Feature ID at geonames.usgs.gov/apex/f?p=gnispq

This will give you the source of the database entry, possibly a list of
alternate names, sometimes a note like "location approximate", and
sometimes the history of the decision process if more than one name
had been proposed for the feature. Also documentation of official
name changes.

One more thing to know about GNIS: entries are never deleted. If a
feature no longer exists, the name gets "(historical)" appended to
it. This may have happened after the feature was imported to OSM,
so it may not show in the OSM database.

Unfortunately the GNIS database is no longer fully maintained due to
budget constraints, so you can't be sure if features still exist even
if they are not flagged as "(historical)".

As to mapping in OSM: I usually remove any "(historical)" feature.
For the others, I improve the location if possible, and if the feature
can be represented as an area, I draw that area/polygon.
Instead of deleting the original POI, I now reuse that node as part
of the outline of the feature and only move the tags to the area, so
someone looking at the object details can notice that one of the nodes
is a lot older than the others and still find the osm history of the
feature on that node.

Wolfgang
( lyx @ OSM )

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

Re: [Talk-us] dubious church node

[Jesse B. Crawford]

I also seem to have observed that, at least in rural New Mexico where I
do most of my mapping, GNIS features like historic places seem to have
only been entered to the resolution of what town they were in, and then
all ended up at something like the centroid of the town or county
limits. The result can be rather odd - before I moved them all to more
correct locations, the parking lot of a particular bank in Socorro, NM
contained over a half dozen historic homes. What an amazing place to
visit!

So this is another thing to keep in mind, it seems like the GNIS
locations may be quite a bit less granular for certain types of
features. The historic features are a good example as they were
presumably all taken from the registry and so they *do* exist, but it
can take a lot of legwork to figure out where.

-- 
Jesse B. Crawford

https://jbcrawford.us
je...@jbcrawford.us
GPG 0x4085BDC1

On Sat, Sep 30, 2017, at 11:05 AM, Mark Wagner wrote:
> On Sat, 30 Sep 2017 15:11:06 +0700
> Dave Swarthout  wrote:
> 
> > "Second, many entries have their coordinates specified using the old
> > NAD 27 datum, but somewhere along the line, that fact was lost and the
> > coordinates were assumed to be in either NAD 83 or WGS 84.  This
> > results in an offset that increases the further you go from central
> > Indiana; the offset in Alaska is upwards of a hundred meters to the
> > west."
> > 
> > Wow, thanks for that. If I understand what you're saying, this means
> > many of the old GNIS nodes will be positioned about 100 meters east
> > of where they should be? Or do I have your statement turned around?
> 
> Depending on where in the process the error was made, it could go
> either way, but in my experience, the nodes are usually positioned too
> far to the east.
> 
> -- 
> Mark
> 
> ___
> 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] dubious church node

[Mark Wagner]

On Sat, 30 Sep 2017 15:11:06 +0700
Dave Swarthout  wrote:

> "Second, many entries have their coordinates specified using the old
> NAD 27 datum, but somewhere along the line, that fact was lost and the
> coordinates were assumed to be in either NAD 83 or WGS 84.  This
> results in an offset that increases the further you go from central
> Indiana; the offset in Alaska is upwards of a hundred meters to the
> west."
> 
> Wow, thanks for that. If I understand what you're saying, this means
> many of the old GNIS nodes will be positioned about 100 meters east
> of where they should be? Or do I have your statement turned around?

Depending on where in the process the error was made, it could go
either way, but in my experience, the nodes are usually positioned too
far to the east.

-- 
Mark

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

Re: [Talk-us] dubious church node

[Brian May]

On 9/29/2017 11:06 PM, Kevin Kenny wrote:
On Fri, Sep 29, 2017 at 9:33 PM, Mark Bradley 
> wrote:



In the course of my mapping in the American Midwest, I have come
across several small country churches of GNIS origin that no
longer exist.  Often there will be a nearby cemetery, but the
church facility is gone.  I simply delete the node.  In one case I
know of, the church building was converted into a home, so I
remapped it accordingly.


Of course, if the cemetery is there on the ground, then it should be 
mapped. But deleting the node for a demolished church is entirely 
appropriate. For a church converted to a private home, consider:


building=detached historic:amenity=place_of_worship historic:name=* etc.

if the building still resembles a church.


For any arm-chair mappers out there, you cannot assume the location of 
the original GNIS point is accurate at all, unless you have up to date 
evidence it is. So if you see a church point sitting on what looks like 
a house in a residential neighborhood on the aerial, then either delete 
it,  mark it as a FIXME or leave it alone. The person working for the 
Feds who originally mapped the point may have been miles off.


A few thoughts:

Churches from GNIS seem to be one of the biggest "map noise" features 
for areas I look at. Sometimes the locational accuracy is spot on, 
church is still there and everything is great. Sometimes the church is a 
mile and half down the road on a different block. Sometimes its in the 
middle of the highway. Sometimes in the water, etc. When I am quickly 
reviewing an area and I see a church point in the water or on a road, I 
usually just move it to a halfway plausible location without doing more 
research. It would be nice to have a fairly solid process for reviewing 
these with external data that is of known high quality.


I did a little playing around with the new USGS Map VIewer [1] and it 
has a Structures layer.  This appears to be part of the volunteer corps 
thing w/ USGS, which was (is?) a national program to provide higher 
accuracy points focused on buildings and structures.  I found this [2] 
from 2012 that provides an overview. Looks like they intended to 
contribute back to OSM - but no word on that in the doc. Found this site 
as well [3], but out of time to dig into it for now. Anyone know more 
about this Structures layer?


In the USGS Map Viewer, you can click on a structure and see details 
about it. Some say source=centroid - to me this means parcel centroid. 
Many have addresses as well. The map viewer allows you to switch the 
base map to OSM. So then you get a nice QA tool to check OSM features in 
an area. The structures layer doesn't include churches, but cemeteries 
are included. Other features include Post Offices, State Capitol 
Buildings, Hospitals / Medical Centers, Police Stations, Prisons, 
Colleges, Technical Schools, Schools, Campgrounds, Trailheads and 
Visitor Information Centers.


I have a statewide parcels layer that just shows church polygons and 
labels that I use sometimes use as well for checking churches - others 
are welcome to use it if interested.


[1] https://viewer.nationalmap.gov/advanced-viewer/
[2] https://pubs.usgs.gov/of/2012/1209/pdf/ofr2012-1209.pdf
[3] https://nationalmap.gov/TheNationalMapCorps/#

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

Re: [Talk-us] dubious church node

[Brian May]

On 9/30/2017 3:19 AM, Mark Wagner wrote:

Second, many entries have their coordinates specified using the old NAD
27 datum, but somewhere along the line, that fact was lost and the
coordinates were assumed to be in either NAD 83 or WGS 84.  This
results in an offset that increases the further you go from central
Indiana; the offset in Alaska is upwards of a hundred meters to the
west.


I found this [1] page that says:.



9. What datum applies to the geographic coordinates in the GNIS Database?
All coordinates in the database are in NAD 83. They were converted from 
NAD 27 in September 2005.




And this page [2] which appears to be official metadata that doesn't 
mention a datum - but it was written in 1994 (see very bottom of page).


[1] https://geonames.usgs.gov/domestic/faqs.htm
[2] https://geonames.usgs.gov/metadata.html

Brian


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

Re: [Talk-us] dubious church node

[Marc Gemis]

> Another point is that if you have the outline for something that GNIS shows
> as a node, please conflate! I've done that with a lot of buildings and parks
> locally - just copy-and-paste the GNIS tags from the node to the polygon and
> then delete the node.

The utilsplugin2 [1] for JOSM and it's replace geometry is another way
to do this quickly.

m.

[1] https://wiki.openstreetmap.org/wiki/JOSM/Plugins/utilsplugin2

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

Re: [Talk-us] dubious church node

[Dave Swarthout]

"Second, many entries have their coordinates specified using the old NAD
27 datum, but somewhere along the line, that fact was lost and the
coordinates were assumed to be in either NAD 83 or WGS 84.  This
results in an offset that increases the further you go from central
Indiana; the offset in Alaska is upwards of a hundred meters to the

west."

Wow, thanks for that. If I understand what you're saying, this means many
of the old GNIS nodes will be positioned about 100 meters east of where
they should be? Or do I have your statement turned around?

The mine whose position I last adjusted, the Case Mine in the Chugach
Mountains on the Kenai Peninsula, was quite a distance from an area of bare
ground (visible only in ESRI) where an old mine site might have been. The
original position was to the east of that bare area. I didn't measure the
distance but will do that next time I come across such a mine.  The bare
spot also happens to be where the USGS Topo places the mine, consequently,
I felt moving it was justified.

I'm also guessing that the other Case Mine node,  the"duplicate" I
mentioned earlier, represents perhaps a second mine_entrance on the same
mining claim. However, there is nothing west of that node to provide any
clue to guide a repositioning, nor does it appear on theUSGS Topo map, so I
left it where it was.

On Sat, Sep 30, 2017 at 2:19 PM, Mark Wagner  wrote:

> On Sat, 30 Sep 2017 06:56:31 +0700
> Dave Swarthout  wrote:
>
> > Glad you mentioned that GNIS import, Ian.
> >
> > This isn't a pressing issue but I've been doing considerable mapping
> > in Alaska and encounter GNIS features constantly. Many of them are
> > nodes and refer to mines, usually abandoned mines, and contain
> > tagging that JOSM complains about, for example, using landuse=quarry
> > on a node. Sometimes I delete that tag and add man_made=mineshaft or
> > similar tagging but it's often not clear if the node is in the proper
> > location. The newer, high-resolution imagery will often suggest a
> > more likely spot for the node, and sometimes I'll move the node
> > there, but usually it isn't obvious. There are also duplicate nodes,
> > that is, mines having the same name but in a slightly different
> > position and carrying a different GNIS reference number.
> >
> > Can you provide some guidance about the accuracy of the positions, the
> > duplication, and perhaps weigh in on possible tagging scenarios?
>
> In my experience, there are two common sources of position error in
> GNIS:
>
> First, many GNIS entries are pulled off of old USGS topo maps.  These
> are of limited resolution, and you can't get a position more accurate
> than about a city block.  It's not much of an error, but when you're
> used to coordinates that will lead you to a specific door, it's
> something to keep in mind.
>
> Second, many entries have their coordinates specified using the old NAD
> 27 datum, but somewhere along the line, that fact was lost and the
> coordinates were assumed to be in either NAD 83 or WGS 84.  This
> results in an offset that increases the further you go from central
> Indiana; the offset in Alaska is upwards of a hundred meters to the
> west.
>
> For churches, hospitals, post offices, and other facilities in towns,
> it's not unusual for them to take the same coordinates as the center of
> the town.  This mis-positioning may be combined with one or both of the
> above.
>
> The other common error you'll encounter is that the tagging is only
> approximate as to type.  This is most obvious with medical facilities:
> everything from doctors' offices to retirement homes gets tagged as
> "amenity=hospital".  More common but less noticeable is that a wide
> range of vaguely recreation-related things get tagged as "leisure=park"
> -- in particular, watch out for historic markers tagged as such.
>
> Your quarries are subject to this same type-approximation: everything
> from a county road department's gravel pit to an extensive complex of
> mineshafts is tagged as "landuse=quarry", as are some mining-related
> industrial facilities.
>
> --
> Mark
>
> ___
> Talk-us mailing list
> Talk-us@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-us
>



-- 
Dave Swarthout
Homer, Alaska
Chiang Mai, Thailand
Travel Blog at http://dswarthout.blogspot.com
___
Talk-us mailing list
Talk-us@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-us

Re: [Talk-us] dubious church node

[Mark Wagner]

On Sat, 30 Sep 2017 06:56:31 +0700
Dave Swarthout  wrote:

> Glad you mentioned that GNIS import, Ian.
> 
> This isn't a pressing issue but I've been doing considerable mapping
> in Alaska and encounter GNIS features constantly. Many of them are
> nodes and refer to mines, usually abandoned mines, and contain
> tagging that JOSM complains about, for example, using landuse=quarry
> on a node. Sometimes I delete that tag and add man_made=mineshaft or
> similar tagging but it's often not clear if the node is in the proper
> location. The newer, high-resolution imagery will often suggest a
> more likely spot for the node, and sometimes I'll move the node
> there, but usually it isn't obvious. There are also duplicate nodes,
> that is, mines having the same name but in a slightly different
> position and carrying a different GNIS reference number.
> 
> Can you provide some guidance about the accuracy of the positions, the
> duplication, and perhaps weigh in on possible tagging scenarios?

In my experience, there are two common sources of position error in
GNIS:

First, many GNIS entries are pulled off of old USGS topo maps.  These
are of limited resolution, and you can't get a position more accurate
than about a city block.  It's not much of an error, but when you're
used to coordinates that will lead you to a specific door, it's
something to keep in mind.

Second, many entries have their coordinates specified using the old NAD
27 datum, but somewhere along the line, that fact was lost and the
coordinates were assumed to be in either NAD 83 or WGS 84.  This
results in an offset that increases the further you go from central
Indiana; the offset in Alaska is upwards of a hundred meters to the
west.

For churches, hospitals, post offices, and other facilities in towns,
it's not unusual for them to take the same coordinates as the center of
the town.  This mis-positioning may be combined with one or both of the
above.

The other common error you'll encounter is that the tagging is only
approximate as to type.  This is most obvious with medical facilities:
everything from doctors' offices to retirement homes gets tagged as
"amenity=hospital".  More common but less noticeable is that a wide
range of vaguely recreation-related things get tagged as "leisure=park"
-- in particular, watch out for historic markers tagged as such.

Your quarries are subject to this same type-approximation: everything
from a county road department's gravel pit to an extensive complex of
mineshafts is tagged as "landuse=quarry", as are some mining-related
industrial facilities.

-- 
Mark

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