2011/2/18 Peter Wendorff <wendo...@uni-paderborn.de>: Hi Peter,
thank you for this explanatory words. > Is not working and "is using wrong semantics" is not the same. It is IMHO not consistent with current tagging (that's why I said it is not working), because building:levels was there long before min_level. Building levels is defined as the numbers of levels a building has. In your example illustration this is 2 levels for the "hovering" part. > You are an architect and from that perspective you are completely right. > My purpose with this design of the tagging scheme was something often > applied in OSM: backwards compatibility. IMHO you didn't achieve this because you are redefining the tag building_levels. > As bridges do not appear alone and instead are always part of a building > including the sides of the bridge, the building as a whole would have been > tagged with building_levels=7. no idea. I would probably have made 3 "buildings" 2 with levels=7 and one with levels=2 (or a footway, bridge=yes). For a very rough "building as a whole"-model it would also be fine to define this as levels=7, but as soon as you are refining this, you would not keep the levels=7 for the part that has only 2 levels. > So far this obviously don't describe the bridge of course. > Now let's think about this bridge as a whole through the building block of 7 > levels height. > And that's what the min_level does: it raises the bottom about the min_level > count - in this example, it raises the bottom about 5 levels. I understood the concept, it is basically a boolean operation (subtraction of volumes). It would require from every user of the data (say you wanted to calculate the gross floor area (BGF in German) which is the sum of the area of all levels including structures, so equal to the size of the osm-polygon multiplied by levels) to understand this concept and to subtract building_levels by min_level in order to get the real building levels. IMHO this introduces unnecessary complexity and will create lots of errors because the concept is not "the standard one". > I played with the alternative more common for you, but what would that mean? > Only being able to interpret building_levels and not supporting min_levels > would lead to false assumptions about the height of the building. please keep in mind, that level heights are not constant. In a theatre or auditorium you will have few levels (typically one main one) but they would be quite high. > A building > with a roof above 7 floors would be interpreted as a building with height of > 2 floors. It _is_ a building (or part) with 2 floors, hanging between 2 other buildings (parts) with 7 levels. > Let's stay at 3D-Rendering as the example application, you would have a > U-shaped instead of a A-shaped building. yes, if you were doing 3D-Rendering and you didn't interpret the base elevation of the building you would get an ugly rendering. So what? If the building was 8 levels of which 4 are underground you also would get a wrong rendering. If the levels had different heights and you don't tag this, you will also get wrong renderings. > yes, but not against interception of most people: Even a building with a > tunnel at the basement level with 6 levels above is seen as a > 7-level-building. Sorry, I don't get that sentence. Can you explain? > And if you want to have the number of levels inside the building you can > simply calculate it by building_levels - min_level. before you added your proposal you simply had to look at the tag building_levels. It was all there. No need to calculate. > The definition of building_levels in the wiki > (http://wiki.openstreetmap.org/wiki/Proposed_features/Building_attributes) > states "Number of stories, including the ground floor". > My interpretation is: There is no explicit definition of the handling of > "missing levels" like "tunnels". I don't get the tunnel-argument (tunnels are IMHO ways that pass through the building, they are not part of the building). Also I don't see "missing levels". Number of stories, including ground floor doesn't leave too much space for ambiguities. The only thing strange is that it focuses on "ground floor" which could (maybe) imply that underground levels are not included. Generally I would expect all (main) levels to be included. "Main" levels would be levels with a height adequate for people to stand/walk (ca. 2 meters, note that this is not according to the German building law, which defines "Vollgeschoss" height > 250 cm (Berlin, there is also other values in other regions) and suitable for people to stay (light, air)). > But it's clearly stated "including ground floor". So I would say: it does > not depend on the existance of a raised first floor. > Even for you as an architect I don't think that you define the first of the > 2 raised floors of the example bridge is a "ground floor". "raised floor" would be something raised above ground (I'd say 1 meter or so, de:"Hochparterre"), in this case I wouldn't speak of a "raised floor", this is a "bridge". A ground floor would be the lowest "floor" (which is not underground). A floor has to have his ceiling at least 140 cm above the surrounding ground in order not to be "underground". Additionally you will have to be able leave the building without additional stairs or help. (Not sure for the precise definition, it doesn't seem to be important for the law, maybe there is exceptions e.g. a level raised 1,2 m might still be considered "ground floor"). I am quite sure the "ground floor" will have to be around ground level --- and I am sure that a floor like in your example 15-20 meters above ground will not be considered "ground floor" ;-) cheers, Martin _______________________________________________ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
