On 09/20/2010 09:50 PM, Hendrik Oesterlin wrote:
So, it is more or less based on luck, whether the calculation of time is better or not. In any case, the time calculated should be longer. Matters could be improved by for example making an automatic split for a road if the segment containing speed-lowering poi gets too long.
This reduces the speed in both directions?
That is another problem with road speed reduction based on unidirectional poi. (unless you make some workaround such as generate overlaying one-way streets for a workaround or whatever).
You can see the effective road speed if you put your Garmin device in "simulation" mode and route to some POI.
Simulation yes. Routing, no. For example, routing for a faster roads ignores mostly minor roads and seems to caltulate faster route based on what was not ignored. So, a major road with lots of traffic_calming and traffic_signals gets preferred. :( -- harri