OSM is a truly incredible resource. It's often lightyears ahead of proprietary maps in terms of feature accuracy. Often far behind in terms of points-of-interest.
It powers painless adding of maps to so many products, painless offline data analytics and joins (I built https://wildpocket.com/ by joining OSM to iNaturalist data).
Truly incredible and under appreciated by general public. I love street complete, it opens up OSM to so many beginners with the simple UX. I think there's an ongoing need for "beginner-friendly" OSM tools - I've prototyped https://mapmender.com/ to help map the "last mile" routes for emergency response that are inherently hard to keep updated. It's a WIP passion project, feedback welcome
A few days ago I was made aware of Street Complete here on HN. Same thing there - it makes it incredibly easy to submit additional data, such as opening hours and road types, to OSM.
Is there any work on tooling for points of interest? Google Maps makes it easy for me to search and update when I see changes e.g. closed for vacation, etc.
> In some cases borders may overlap without making it an incorrect tagging. It may be result of agreement between countries to administer area together or to mark areas not controlled by any country, with both claiming it.
Fun fact: when you think of boundary disputes, two things tend to be assumed. One, this only happens in 'certain' countries. Two, this always means more than one country claims the same piece of land.
As for two: because of complex reasons to with their eventual bullshit bargaining position, there are several slivers of land wedged between the respectively claimed borders of Serbia and Croatia which belong to… nobody¹ (https://www.openstreetmap.org/relation/19478817). Although the local Serbian border patrols will gladly point their rifles in your general direction to point out that this does not, in fact, mean you get to be there as far as they're concerned.
1: Head-line grabbing claims by assorted dreamers notwithstanding; hence Verdis.
My current favorite boundary dispute is between California and Nevada.
California entered the union in 1850 and it's north-eastern boundary was defined a specific way(120 degrees west of Greenwich), Nevada entered 14 years later and it's western boundary was defined with a different method(43 degrees west of Washington), at the time it was thought these two methods were equivalent, but this was later proved otherwise. It appears the main strip got resolved but there is still a weird triangular strip of land on the diagonal claimed by both sides. There were also a number of surveying mistakes, but the root of the problem appears to be the accepted but incompatible definitions. It turns out that congress has no authority to change state borders once they are set(Somehow this both surprised me and made total sense at the same time), so the states themselves have to resolve it. There is currently an "agreement" on how to interpret the border, but neither state has changed it's constitution so the dispute is still technically ongoing.
Wait, so it was actually believed that the Washington meridian just happened to fall exactly on an integer number of degrees (120-43 =77) west of Greenwich?
It is close, and if you consider Washington as a district there is a meridian exactly 77 degrees west somewhere in there. But I am here eyeballing maps and if the "Washington meridian" is considered the point of the capital it is ~ 1000 feet west of -77.0, should have chosen Stanton park. Seeing as Washington was a planned city I wonder if the capitol point was supposed to be -77.0 west but limitations in surveying accuracy put it where it is. A thousand foot error is not bad for 1800. In fact the very public line at Greenwich is off by ~ 50 feet.
On the subject of eyeballing maps are there any online openstreetmap derivatives that have better cartography tools? (crosshair, lat long at specific point, distance and bearing lines, etc) I probably could install qgis for it, and osmand has a bunch of these tools. but the default osm page sort of sucks here.
Reading the constitutions of both states in order to pin down exactly where the problem is, the Wikipedia article is sort of vague. The Nevada document specifically references "thence in a North Westerly direction along the Eastern boundary line of the State of California to the forty-third degree of Longitude West from Washington" So that is not the problem, the border is whatever California says it is. It must be a north-south strip in the difference between "the forty-third degree of Longitude West from Washington" and "Running south on the line of said 120th degree of West Longitude"
The funny thing I found in the Nevada constitution was
"All territory lying West of and adjoining the boundary line herein prescribed, which the State of California may relinquish to the Territory or State of Nevada, shall thereupon be embraced within and constitute a part of this State."
We reluctantly agree to your existing borders, but we really wanted the Sierra Nevada ridgeline to be the border and if you want to give it to us we are going to put it right in the constitution to accept it. The passive aggressive tone in such a formal document just kills me.
As an excellent example of #2 we've got Bir Tawil and the Halaib Triangle - both exist along the Sudanese-Egyptian border. Two different historical border drawings have been presented which each argue that the Halaib Triangle belongs to a different country and the triangle has 27k inhabitants. Funnily enough, each map that yields that valuable chunk of land to each country actually yields that Bir Tawil should be possessed by the other country (and Bir Tawil is comparatively worthless). As a result neither country claims Bir Tawil since it could be seen as an admission that the other country has a rightful claim to Halaib Triangle.
Usually these weird chunks of land that lay unclaimed are usually the result of agreements or contests around more valuable chunks of land that cause voids in the border - sometimes just due to historical oversight.
For Myanmar, having "De facto independent areas (e.g. Wa state)" at admin_level=3 is wildly ambitious, since not only are these boundaries constantly shifting slightly due to the ongoing civil war, the actors involved don't even necessarily have the state capacity to accurately demarcate their own territorial claims.
Documenting disputed borders or borders in war zones seems tricky in general. Not only can they be constantly shifting, but even at a fixed point in time, the boundary is probably wildly different depending on which of the belligerent parties you may ask - or in extreme cases, a party may deny the demarcation exists at all.
Disputes are not such a big problem, since you can have multiple independent versions of the boundary with claimed_by= and disputed_by= tags differentiating them. But you do need to find out somehow where the claimed boundary even is. Like, does the Wa State have an Office of Cartography where you can ask for the latest official maps? I haven't looked super hard, but I doubt it exists.
Nice to see someone is keeping track of all of this in a logical way for all countries.
> Note: It has been decided not to add administrative boundaries of [super-national administrations] in OpenStreetMap. So please don't use admin_level=1.
Damn, I was hoping to see groups of countries (admin_level=1), and then groups of planets (admin_level=0), and then groups of stars (admin_level=-1), and then groups of galaxies (admin_level=-2). L. Ron Hubbard has surely established governments at all of these levels!
OSM is a truly incredible resource. It's often lightyears ahead of proprietary maps in terms of feature accuracy. Often far behind in terms of points-of-interest.
It powers painless adding of maps to so many products, painless offline data analytics and joins (I built https://wildpocket.com/ by joining OSM to iNaturalist data).
Truly incredible and under appreciated by general public. I love street complete, it opens up OSM to so many beginners with the simple UX. I think there's an ongoing need for "beginner-friendly" OSM tools - I've prototyped https://mapmender.com/ to help map the "last mile" routes for emergency response that are inherently hard to keep updated. It's a WIP passion project, feedback welcome
A few days ago I was made aware of Street Complete here on HN. Same thing there - it makes it incredibly easy to submit additional data, such as opening hours and road types, to OSM.
https://f-droid.org/packages/de.westnordost.streetcomplete
Is there any work on tooling for points of interest? Google Maps makes it easy for me to search and update when I see changes e.g. closed for vacation, etc.
I don't have a specific reference, but here is a useful catalog of OSM-powered/OSM-editing tools:
https://wiki.openstreetmap.org/wiki/List_of_OSM-based_servic...
https://every-door.app is my favorite for light-weight POI editing. You can blast through an entire shopping mall very quickly
> In some cases borders may overlap without making it an incorrect tagging. It may be result of agreement between countries to administer area together or to mark areas not controlled by any country, with both claiming it.
Fun fact: when you think of boundary disputes, two things tend to be assumed. One, this only happens in 'certain' countries. Two, this always means more than one country claims the same piece of land.
As for one: even the Netherlands and Germany claim the same bit of sea. Boundary status: unresolved (https://www.openstreetmap.org/relation/19478817#map=12/53.56...), but don't worry, we're friends and allies these days.
As for two: because of complex reasons to with their eventual bullshit bargaining position, there are several slivers of land wedged between the respectively claimed borders of Serbia and Croatia which belong to… nobody¹ (https://www.openstreetmap.org/relation/19478817). Although the local Serbian border patrols will gladly point their rifles in your general direction to point out that this does not, in fact, mean you get to be there as far as they're concerned.
1: Head-line grabbing claims by assorted dreamers notwithstanding; hence Verdis.
My current favorite boundary dispute is between California and Nevada.
California entered the union in 1850 and it's north-eastern boundary was defined a specific way(120 degrees west of Greenwich), Nevada entered 14 years later and it's western boundary was defined with a different method(43 degrees west of Washington), at the time it was thought these two methods were equivalent, but this was later proved otherwise. It appears the main strip got resolved but there is still a weird triangular strip of land on the diagonal claimed by both sides. There were also a number of surveying mistakes, but the root of the problem appears to be the accepted but incompatible definitions. It turns out that congress has no authority to change state borders once they are set(Somehow this both surprised me and made total sense at the same time), so the states themselves have to resolve it. There is currently an "agreement" on how to interpret the border, but neither state has changed it's constitution so the dispute is still technically ongoing.
https://en.wikipedia.org/wiki/Lake_Tahoe#Interstate_boundary...
Wait, so it was actually believed that the Washington meridian just happened to fall exactly on an integer number of degrees (120-43 =77) west of Greenwich?
It is close, and if you consider Washington as a district there is a meridian exactly 77 degrees west somewhere in there. But I am here eyeballing maps and if the "Washington meridian" is considered the point of the capital it is ~ 1000 feet west of -77.0, should have chosen Stanton park. Seeing as Washington was a planned city I wonder if the capitol point was supposed to be -77.0 west but limitations in surveying accuracy put it where it is. A thousand foot error is not bad for 1800. In fact the very public line at Greenwich is off by ~ 50 feet.
On the subject of eyeballing maps are there any online openstreetmap derivatives that have better cartography tools? (crosshair, lat long at specific point, distance and bearing lines, etc) I probably could install qgis for it, and osmand has a bunch of these tools. but the default osm page sort of sucks here.
Presumably defined rather than just happened, but not surveyed exactly enough. The US Capitol building is at 77.009°W.
Reading the constitutions of both states in order to pin down exactly where the problem is, the Wikipedia article is sort of vague. The Nevada document specifically references "thence in a North Westerly direction along the Eastern boundary line of the State of California to the forty-third degree of Longitude West from Washington" So that is not the problem, the border is whatever California says it is. It must be a north-south strip in the difference between "the forty-third degree of Longitude West from Washington" and "Running south on the line of said 120th degree of West Longitude"
The latest California constitution just says to reference the original 1849 for the boundaries https://archives.cdn.sos.ca.gov/pdf/1849-california-constitu...
Nevada https://www.leg.state.nv.us/Const/NVConst.html#Art14 amended in 1982, but the original 1864 document has the same geographic description with a different starting point for some reason.
The funny thing I found in the Nevada constitution was
"All territory lying West of and adjoining the boundary line herein prescribed, which the State of California may relinquish to the Territory or State of Nevada, shall thereupon be embraced within and constitute a part of this State."
We reluctantly agree to your existing borders, but we really wanted the Sierra Nevada ridgeline to be the border and if you want to give it to us we are going to put it right in the constitution to accept it. The passive aggressive tone in such a formal document just kills me.
As an excellent example of #2 we've got Bir Tawil and the Halaib Triangle - both exist along the Sudanese-Egyptian border. Two different historical border drawings have been presented which each argue that the Halaib Triangle belongs to a different country and the triangle has 27k inhabitants. Funnily enough, each map that yields that valuable chunk of land to each country actually yields that Bir Tawil should be possessed by the other country (and Bir Tawil is comparatively worthless). As a result neither country claims Bir Tawil since it could be seen as an admission that the other country has a rightful claim to Halaib Triangle.
Usually these weird chunks of land that lay unclaimed are usually the result of agreements or contests around more valuable chunks of land that cause voids in the border - sometimes just due to historical oversight.
For Myanmar, having "De facto independent areas (e.g. Wa state)" at admin_level=3 is wildly ambitious, since not only are these boundaries constantly shifting slightly due to the ongoing civil war, the actors involved don't even necessarily have the state capacity to accurately demarcate their own territorial claims.
The Wa State in particular appears to have undergone some "boundary fixing" a few weeks ago https://www.openstreetmap.org/relation/9399863 but whether it now accurately reflects realities on the ground is anyone's guess. I'm inclined to believe the Wikimedia map https://commons.wikimedia.org/wiki/File:Myanmar_civil_war.sv... showing Tangyan in Wa State territory is more up-to-date, unless they withdrew for some reason without it making the news. (The most recent I could find is this article from March where they were mining quartz in the area: https://english.shannews.org/archives/29385 )
Documenting disputed borders or borders in war zones seems tricky in general. Not only can they be constantly shifting, but even at a fixed point in time, the boundary is probably wildly different depending on which of the belligerent parties you may ask - or in extreme cases, a party may deny the demarcation exists at all.
Disputes are not such a big problem, since you can have multiple independent versions of the boundary with claimed_by= and disputed_by= tags differentiating them. But you do need to find out somehow where the claimed boundary even is. Like, does the Wa State have an Office of Cartography where you can ask for the latest official maps? I haven't looked super hard, but I doubt it exists.
I had just watched a fantastic video about this very topic as it applies to the US https://www.youtube.com/watch?v=b5JbDuf--ng
Nice to see someone is keeping track of all of this in a logical way for all countries.
> Note: It has been decided not to add administrative boundaries of [super-national administrations] in OpenStreetMap. So please don't use admin_level=1.
Damn, I was hoping to see groups of countries (admin_level=1), and then groups of planets (admin_level=0), and then groups of stars (admin_level=-1), and then groups of galaxies (admin_level=-2). L. Ron Hubbard has surely established governments at all of these levels!
I started to contribute a lot to Openstreetmap since someone submitted the StreetComplete app a while ago !
It's amazing that something like this exists.
TIL of 'townland' - https://en.wikipedia.org/wiki/Townland - a subdivision mostly in Ireland and Scotland. (While I looked up UK)
Speaking of which, something odd seems to be happening in Peru ocean limits https://osm.org/go/NSkYk-
Fittingly, Peru co-invented the 200-mile zone and still hasn't ratified UNCLOS — it uses its own dominio marítimo concept instead of an EEZ.
Looks like this is a known ongoing dispute: https://wiki.openstreetmap.org/wiki/Disputed_territories#Sou...