On Fri, 2015-03-27 at 14:07 +0000, Jan van Bekkum wrote: > It is a bit of a philosophical question: do you prefer a placeholder > or a polygon of which you don't know how correct it is, for example a > forest behind the campsite that may or may not be part of the > campground.
In natural surroundings, a place holder node is the lesser of two evils. To mark an arbitrary poly implies camping within is somehow better than outside. And that is totally misleading. But a single node cannot carry the #4 model associating different features or constraints. So a very little polygon ? One that implies its a 'pitch' rather than a campground ? It says "I camped right here". David > > > On Fri, Mar 27, 2015 at 1:57 PM Marc Gemis <marc.ge...@gmail.com> > wrote: > In many cases you will be able to determine the area from the > aerial images (thinking of Western European campsites). > I assume that in the campsites you visited, the actual area > was rather fuzzy and that the exact area will never been > known, not ? OSM has no solution for fuzzy areas anyhow. > > > Is it difficult to obtain an approximation of the area when > you already go through the effort to position all the > amenities as individual nodes ? > you can always leave a note or fixme tag to indicate that the > shape has to be established. > > > just my .5 cents > > > regards > > > m > > On Fri, Mar 27, 2015 at 12:37 PM, Jan van Bekkum > <jan.vanbek...@gmail.com> wrote: > So if you don't know the real shape of the polygon it > would be best to create a placeholder polygon (like a > circle - it will be clear that it is a placeholder) > and put all amenities inside it until the real shape > is known. > > > On Fri, Mar 27, 2015 at 10:33 AM Marc Gemis > <marc.ge...@gmail.com> wrote: > > Overpass understands this. When I look for all > toilets in the "Zoo Antwerpen" with [1], I > only find toilets in that Zoo > > > regards > > > m > > > [1] http://overpass-turbo.eu/s/8qL > > On Fri, Mar 27, 2015 at 10:24 AM, Marc Gemis > <marc.ge...@gmail.com> wrote: > > OK, I did not know that ! Is > this "is-in-polygon" test > something that > is already being done ? > Examples ? > > > Nominatim that adds the address of the > building to the POI is an example of a > similar test / algorithm. > Sorry, don't know any other examples. > But it just makes sense that you do > not have to define inclusion of > something when you can determine that > from it's position. > > > I also only know 1 website that > supports the site relation, the > geschichtskarte for historical items > > > regards > > > m > > > > > _______________________________________________ > Tagging mailing list > Tagging@openstreetmap.org > https://lists.openstreetmap.org/listinfo/tagging > > _______________________________________________ > Tagging mailing list > Tagging@openstreetmap.org > https://lists.openstreetmap.org/listinfo/tagging > > > > _______________________________________________ > Tagging mailing list > Tagging@openstreetmap.org > https://lists.openstreetmap.org/listinfo/tagging > _______________________________________________ > Tagging mailing list > Tagging@openstreetmap.org > https://lists.openstreetmap.org/listinfo/tagging _______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
