On Sat, 1 Aug 2020 at 07:21, Paul Norman via Tagging < [email protected]> wrote:
> On 2020-07-31 8:21 a.m., Andy Townsend wrote: > > On 26/05/2020 00:20, Alan Mackie wrote: > > Has this edit war stabilised? > > Apparently it has been blocking coastline updates across the whole world > for *months *now. > > https://osmdata.openstreetmap.de/data/land-polygons.html > https://github.com/fossgis/osmdata/issues/7 > > (picking this thread up again because there still hasn't exactly been a > meeting of minds here) > > land polygons have been generated (see > https://osmdata.openstreetmap.de/data/land-polygons.html ) and > https://github.com/fossgis/osmdata/issues/7 has been resolved by manually > "releasing" the coastline. The current situation in OSM is > https://overpass-turbo.eu/s/WD8 - at the time of writing this the > coastline crosses the river north of Buenos Aires. > > However, edits are continuing (see > https://www.openstreetmap.org/changeset/88787419 ). I'm not convinced > that moving to one of two extremes, even a small amount at a time, is a > good idea until there's actually been discussion between the proponents of > the various positions. > > For what it's worth, neither extreme position looks the best answer to me > - looking at the salinity change between river to ocean at > https://www.sciencedirect.com/science/article/pii/S0307904X07000716 (see > https://www.sciencedirect.com/science/article/pii/S0307904X07000716 for > the key picture) and looking at > https://de.wikipedia.org/wiki/Datei:Rio_de_la_Plata_BA_2.JPG suggests a > location some way between the two. Despite the NASA photo it looks like > there isn't a "step change" in salinity - and of course values will > fluctuate based on winds and tides etc > > > I live near the coast and have done coastline processing, including a > great deal worldwide during the redaction. > > Salinity and territorial control have seldom been considerations in where > the break between water mapped as waterway=riverbank and natural=coastline > that I have seen. The break is chosen as a convenient place for mappers and > a common view of where the coast of the ocean is, not based on scientific > salinity criteria. For territorial control, look at all the inlets along > the BC or Norwegian coasts. > Perhaps I am an overly literal follower of the wiki, but I had always assumed the coastline should continue inland as far as the tide continues to be noticeable. Mediterranean mapping might be an issue, but elsewhere I think this is fairly clear? If the water is fresh or the waterway still appears to be a river, canal etc, then it seems reasonable that they should also have those tags as well. The coastline and riverbank tags aren't fighting for a common key, so it's not a direct tagging conflict. As for territorial control, there are archipelagic states with territorial waters despite large gaps between all their islands. I'm not sure why inlets or bays pose a problem?
_______________________________________________ Tagging mailing list [email protected] https://lists.openstreetmap.org/listinfo/tagging
