doctorkb wrote:This sounds like a known limitation (i.e. not an "issue", but works-as-designed).

The issue is that, as part of its routing sanity check, Waze won't provide a route that is X times longer than the distance between two points as the crow flies.

...

So this issues has been brought up before and has been determined to be " as-designed"? Probably more so with your Maui example, these are properly designed routes, some used more then others. But at the same time, I do understand the premise of Waze is to navigate commuter traffic and not so much point to point navigation.

With that being said, doesn't sounds like will be be addressed anytime soon, if ever... What would be the proper response to the user? While they want to expand the user base of Waze, saying that your valid, legal route that does not work in Waze because it was "designed" not to, doesn't say much of the service. And don't feel a "you'll have to start at point A.5 vs Point A, doesn't seem right either.

This sounds like a known limitation (i.e. not an "issue", but works-as-designed).

The issue is that, as part of its routing sanity check, Waze won't provide a route that is X times longer than the distance between two points as the crow flies.

This was identified several years ago - routing from Wailea, HI to Hana, HI would fail, but you could route from Wailea to Kahului and that same point in Kahului to Hana just fine.

I don't have a good answer for the user.

All I know is that it was brought up with Waze staff a couple years ago, they confirmed the behaviour and seemed unwilling to adjust it. They suggested, as you did, that we simply instruct the user to plan a route to "A.5" then from "A.5" to "B" upon arrival.

Also, FYI, the route on Maui is actually "properly designed" - the road to Hana is a common tourist thing to do, and many tourists stay in the Wailea area.

I believe the justification is to make up for the potential that a more direct path is just "broken" - can you imagine the backlash if Waze took you 100 miles out of the way simply because a segment was disconnected?

Catching up on other forums. Has this been elevated and/or resolved?

Looks like you're bumping into the limits set on the search algorithm. Notice that if you originate from inside this little peninsula it fails:

https://www.waze.com/editor/?env=usa&lo ... s=59455259

but if you go even further away to the east it succeeds:

https://www.waze.com/editor/?env=usa&lo ... s=59619151

I suspect it has to do with how close you are to start from your destination (as the crow flies) that determines how far the search will be allowed to search in all directions (eg. backwards) to find a route. How far away you can get is most likely affected as well by the angle that you are going away from your destination. The greater the angle, the less you want to go that way.

Not known to me. I suggest to take the discussion to the Navigation and Routing forum.

Always room for improvement. Gmaps works in this case. The sanity test is giving up too quickly just because the points are "close". It's obviously not a question of computing resources since longer routes that manage to circle the lake and pass through the bad destination can work.

There are some soft turns along that route. Perhaps getting rid of them will help? I'd also be interested to see if anything changes with the placement of RPPs for these addresses.

Could there be some sort of error thrown because these points are so close together but yet the route needed to get to them is much longer? I say this as the route seems to fail in either direction using multiple different roads in the area just after a route distance of 50 km. It also seems that I can route to other locations over the exact same roads without any issues. I get issues when I move the start and end points too close together on different sides of the lake.

I note this post which refers to a similar problem related to a lake and "close" start/end points.

viewtopic.php?p=1568189#p1568189