Just some notes on waterway=river vs natural=water+water=river. For me, a riverbank is either a line (a wall) separating the land from the river, or a sloped land area with water-loving plants that can be inundated when the water level rises.
When you describe the water of the river, you describe the river, not the riverbank. And a riverbank isn't a waterway either (you navigate the river, not the bank). It even causes some problems, like when you want to measure the total length of waterways in an area, you have to exclude this strange case. So I would prefer to just rename all waterway=riverbank tags to natural=water+water=river. But sadly, the term riverbank has also been used for canals and streams (which sounds totally bonkers to me), so such a conversion isn't possible when done manually. The page describing the water=* tags also says nothing about using multi polygons. It's just as fine to use natural=water on river segments. So in short, I don't get why there are people who want to keep that awkwardly named tag . Regards, Sander Op 29-jul.-2016 00:09 schreef "Ruben Maes" <ru...@janmaes.com>: > Now back on topic: we have the Leie's banks back on the render. > > On donderdag 28 juli 2016 17:54 Jakka wrote: > > > https://www.openstreetmap.org/note/646584#map=17/50.82877/3.25798&layers=N > > > > Please feedback what and where was the cause > > There were two places where there were problems with relation > http://osm.org/relation/2393380: > > * http://osm.org/relation/2393380#map=17/50.89599/3.34545 > The sluice was a complete mess. It took me some time to figure out what > was going on exactly here. > > * http://osm.org/changeset/41095237#map=19/50.92122/3.42972 > There was a random tiny, excess way in the relation. > > The waterway=riverbank tags were non-uniformly applied. Not all ways had > them, some had an additional random area=yes or area=no. Some had > natural=riverbank, waterway=canalbank or some other weird derivation I had > never heard of. > I stripped those completely. No more waterway=riverbank on this part of > the Leie [note 1]. But if some day consensus is reached (not that that will > ever happen) that the multipolygon approach with {natural=water, > water=river} isn't good after all, it wouldn't be so hard to convert it. > > [note 1] Sorry Glenn, I really couldn't be bothered any more. With a > relation, I can tell JOSM to fetch all member ways, use webtools like the > one you mentioned to check their integrity, and the JOSM validator will > complain if I'm messing up. > > -- > Dit bericht is ondertekend met OpenPGP. > _______________________________________________ > Talk-be mailing list > Talk-be@openstreetmap.org > https://lists.openstreetmap.org/listinfo/talk-be > >
_______________________________________________ Talk-be mailing list Talk-be@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-be