On Sun, 27 Mar 2016 19:16:42 +0200 Anders Fougner <anders.foug...@gmail.com> wrote:
> > > Den 27. mars 2016 19.00.18 CEST, skrev Mateusz Konieczny > <matkoni...@gmail.com>: > >On Sun, 27 Mar 2016 09:50:21 -0700 > >Clifford Snow <cliff...@snowandsnow.us> wrote: > > > >> On Sun, Mar 27, 2016 at 9:18 AM, Martin Koppenhoefer > >> <dieterdre...@gmail.com > >> > wrote: > >> > >> > I agree that a rough polygon seems better than a node because it > >> > allows to estimate the size (a new relation datatype would even > >> > be better, like a collection of (existing/already mapped) things > >> > inside (role) and outside (role) that would serve the same > >> > purpose but make it clear that it is only an estimate / that > >> > there aren't clear borders anyway). > >> > > >> > I don't like boundary=informal though. It should be something > >> > more verbose regarding what kind of region this is > >> > (natural/geographic, (low) mountain range, area of lakes, > >> > forest, desert, plains, cultural, ethnographic, wine, etc.) > >> > > >> > >> A while back one of the conversations on the mailing list was about > >> adding neighborhood boundaries. There was a lot of concern that > >> many neighborhood boundaries are not clearly define which would > >> result in boundary disputes. How is adding a rough boundary for an > >> informal area any different? > >> > >> Worse, if we start adding informal boundaries I can see someone > >> wanting to add the Cascadia [1] (Independance Movement) boundary. > >> > >> > >> > >> [1] > >https://en.wikipedia.org/wiki/Cascadia_%28independence_movement%29 > >> > >> Clifford > >> > > > >Given that idea of tagging natural=bay as polygons is controversial I > >am not expecting this to be a good idea. > > > >Areas with completely undefined borders should not be stored in OSM. > > If I understand correctly what you mean (?), I completely disagree. > Maps are really useless if they cannot contain names of such features > that are hard to define the borders of (bays, cliffs, seas, > neighborhoods, etc.). Where's the border of e.g. the North Sea, the > Bay/Gulf of Bothnia, and Skagerrak, and why should they not be on the > map? (they are probably indicated on every useful map of northern > Europe or Scandinavia) I never said that these should not appear on maps. It is about unfortunate fact such objects have no clear definitions - I had recent unfortunate case of somebody expecting clear definition of borders of Carpathians. In that case answer strongly depends on who and how defines Carpathians. For examples see https://en.wikipedia.org/wiki/Carpathian_Mountains with articles in other languages. Even basic facts like length and list of countries where this mountain range is located are different. OSM is not a proper place to store all possible definitions of "border of Carpathians" or pretend to have the perfect one. _______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging