Hi all,
Sorry to write this in English, but it will take me to much effort to do this in German. As there is much discussion in this thread about the usefulness and possibility of area:highway=x rendering, I thought I would share some of the latest development work on my ArcGIS Renderer for OpenStreetMap (not yet publicly available).
It shows two example areas of the city of Szczecin, Poland, located near the Poland/Germany border, and for the second image with the large flyovers an extra Google Maps comparison.
The rendering shows area:highway features tagged and rendered with surface=x, which means the rendering will show differentiation based on surface, e.g. “asphalt” or “sett”. Unfortunately, since most of the area:highway features tagged and digitized in this area are main roads consisting of asphalt, you won’t see much differentiation, except the small areas forming the “Mazurska” residential road in the lower right part of the first image. Besides rendering of area:highway features, you can also see the renderer fully resolving man_made=bridge as areas based on the layer=x tag in the second image of the large road junction near the Oder / Odra river. Based on test-renderings of the area, I discovered issues with the layer=x tagging of the road lines and road areas and bridges at this large junction, and actually went in to correct this using JOSM, allowing you to see the this correctly resolved.
The rendering of area:highway only kicks in at high zoom and large scale, before that, you just see the normal road classification. I have chosen to display surface over road classification, as I think it adds something new at large scales (>1:2500), and at these large scales and highly zoomed in, road classification starts to become less important. Thus, showing surface is really nice I think.
There is one exception to the surface=x based rendering: area:highway features tagged as area:highway=cycleway are rendered as an overlay layer together with “emergency” road markings in a flat muted red color. This is because I think clearly showing them as cycleways is far more important than knowing the actual surface. Most of the area:highway=cycleway features added as areas, will be in cities anyway, and paved by default in some way, so “surface” is less important than the fact that you can actually easily spot them on the map by their reddish color.
Anyway, I noticed a lot of area:highway features currently lack a surface=x. I know there has been written about the possibility of automatically deriving this information from the corresponding line element, but frankly, knowing this would entail global scale geoprocessing operations on millions upon millions of features, I think this is unrealistic, and then I am not even considering the problems with way and area features not exactly matching at end points / nodes.
I therefore would like to encourage everybody to at least, as a minimum, to additionally add surface=x to area:highway features.
Marco Boeringa
The Netherlands
