> The big problem (as I see it) is that OsmAnd can't calculate long routes.
> The answer is: "You have to add waypoints", and that is also what OsmAnd
> suggests.
Reminds me that I wonder if OsmAnd (or other routers) use some form of
"planning": rather than do a recursive search from start to end, they
first pick some intermediate waypoints.
E.g. I imagine that for a long trip you could do:
- Pick the point right in the middle between start and end
- Look for the closest highway (not in the OSM sense but in the sense
of an important road) and use that as a tentative intermediate waypoint.
- Compute the two halves of the trip.
- Since the tentative intermediate waypoint was chosen somewhat
arbitrarily, it's probably not a good choice. So: remove that
tentative waypoint and instead take the two middle points of the two
halves as new non-tentative waypoints (which are presumably
approximately 25% and 75% along the way, so we only need to recompute
the middle portion).
This way we have to compute 3 half-trips, which is good if computing the
total trip is much longer than computing half of it. Of course, it's
bad if computing the total trip is not much more than twice as long as
computing half the trip.
Also this strategy will likely work poorly when the final trip has to go
around a body of water (so the tentative waypoint may be unreachable,
for example).
Stefan
--
You received this message because you are subscribed to the Google Groups
"OsmAnd" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/osmand/jwvk13rvsno.fsf-monnier%2BInbox%40gnu.org.
