> > Jaak Laineste wrote: >> Also it is very easy to give a route >> which is near optimal: e.g. take a small detour from any crossing, this >> makes alternative routing to a very tricky task. > > See the part of my mail where I mention selecting an alternative which > is at least 50% different to the original route.
Yes, I noticed it. Anyway, when you think about it then this kind of target for an algorithm is way too ambiguous (with too many possibilities) to give reasonable result? This extra penalty for old route approach would give one specific new route, which is optimal in the new graph. _______________________________________________ Routing mailing list [email protected] http://lists.openstreetmap.org/listinfo/routing
