On Mon, 23 Apr 2012 22:03:05 Jason Cunningham wrote: > Problem with that, and why I am said this is far more complex than I first > thought, is some people responding to lanes=1.5 by saying 'computers' only > like whole numbers. This suggests width=4.3 would need to be rounded to > either width=4 or width=5 neither of which would help with solving the > lanes=1.5 problem, because 4m is to narrow for two 2.010m cars, and 5m > arguably doesn't require you to significantly slow down.
No, lanes=* is an integer and width=* is a real number. Furthermore, as an integer, lanes=* can be counted, so it's verifiable, and not subject to someone's guess, or dependent on the type of car they were driving. width=* is measured in metres, and can be fractional. It can be estimated, but everyone makes their estimate based on a standard unit of measurement. Later, the measurement can be refined, but the refinement doesn't depend on which car you are driving. If we record the number of physically painted lanes, or an assumption based on whether the road is oneway or not, and we record the width, then anyone can download the data and figure out if the road is wide enough for *their* vehicle[1]. And really, if a road is 4.3 m wide it doesn't matter how wide your car is, you should probably slow down. Best wishes, Andrew [1] However, you don't know the width of the oncoming vehicles. _______________________________________________ Tagging mailing list [email protected] http://lists.openstreetmap.org/listinfo/tagging
