
Hi Paul,
I don't know when this happened as I haven't tried long distance routing for a while but I noticed today that this is vastly improved. I (amongst others) have commented previously about the sometimes strange routes chosen when trying to route long distances and Mark suggested at the time that we may simply be asking too much of the devices. These routes would work but would be far from optimal i.e from Manchester to Cornwall would go through the middle of Manchester, on to the M6, off the M6 , through Birmingham, on to the M5, off the M5 through Bristol, onto the M5 again and then finally on to the destination.
Today, however, when trying a similar route the given route was M60, M6, M5 with no deviations which to me is exactly the way I would travel. Additionally my 605 can now calculate routes to places further away than it previously could.
As this is such an improvement and I missed it at the time I thought I'd drop a quick line to say thanks to whoever found the magic key.
Thanks for the report, it's great that the routing is working well over long distances. I can't think of any obvious changes apart from the addition of the --remove-short-arcs stuff which would improve the routing if the maps contain short arcs. Oh yes, and there was also a fix to avoid problems with long arcs! So perhaps, that's what has made the difference. If only we could get a grip on the .MDR file format, it would make search work as well as the routing. Cheers, Mark