Unexpected routing change in 3.54 (offline map related, possibly)
Quote from Barkinmadd on 02.03.2024, 08:04I've been using 3.53 to plan a route from Kitchener, Ontario, Canada to the Rock 'n Roll Hale of Fame in Cleveland (start and end route) using 'Fastest' but avoiding highways and unpaved roads. I had one stop along the way.
I'm new to Scenic so I made several copies of the basic route just to play with adding stops and way-points, etc. but invariably I could remove the stops, recalculate the 'Fastest' avoiding highways and unpaved roads and I would get the same route as I did originally. I also have offline maps for the province and states that are included in this route (again, partly for learning the workflow, but will be useful when I actually make the ride).
I played with this same route yesterday (before the 3.54 update) and the route remained the same taking me along a nice stretch of Lake Erie relatively close to the lake.
All of the above is just to provide context for the unexpected behaviour I am reporting below. Sorry for the TLDR.
This morning I update the app to 3.54 (on both the iPad and my iPhone) and then went into the route to edit some points. As I entered the edit mode, Scenic recalculated the route and to my surprised, decided to deviate far south of Lake Erie at one point - a totally different route. I checked that 'Fastest', avoiding highways and unpaved roads was still selected (and it was). This new routing occurred on all my copies of the route on both my iPad and iPhone (I cancelled out of the edit mode each time to retain the original route).
At this point I was a bit confused and concerned. I then noticed in my profile that there was a "map updates available" notification. I thought to myself that perhaps 3.54 needed new offline maps for some reason (maybe this was even documented in the update and I missed it). After updating my offline maps, I returned to each of these routes on both iPad and iPhone and all routes were recalculated to the original route without the southern deviation away from the lake.
I'm not complaining, just reporting. My guess is that if I didn't have the offline maps I would never have noticed a change from 3.53 to 3.54. Cheers.
I've been using 3.53 to plan a route from Kitchener, Ontario, Canada to the Rock 'n Roll Hale of Fame in Cleveland (start and end route) using 'Fastest' but avoiding highways and unpaved roads. I had one stop along the way.
I'm new to Scenic so I made several copies of the basic route just to play with adding stops and way-points, etc. but invariably I could remove the stops, recalculate the 'Fastest' avoiding highways and unpaved roads and I would get the same route as I did originally. I also have offline maps for the province and states that are included in this route (again, partly for learning the workflow, but will be useful when I actually make the ride).
I played with this same route yesterday (before the 3.54 update) and the route remained the same taking me along a nice stretch of Lake Erie relatively close to the lake.
All of the above is just to provide context for the unexpected behaviour I am reporting below. Sorry for the TLDR.
This morning I update the app to 3.54 (on both the iPad and my iPhone) and then went into the route to edit some points. As I entered the edit mode, Scenic recalculated the route and to my surprised, decided to deviate far south of Lake Erie at one point - a totally different route. I checked that 'Fastest', avoiding highways and unpaved roads was still selected (and it was). This new routing occurred on all my copies of the route on both my iPad and iPhone (I cancelled out of the edit mode each time to retain the original route).
At this point I was a bit confused and concerned. I then noticed in my profile that there was a "map updates available" notification. I thought to myself that perhaps 3.54 needed new offline maps for some reason (maybe this was even documented in the update and I missed it). After updating my offline maps, I returned to each of these routes on both iPad and iPhone and all routes were recalculated to the original route without the southern deviation away from the lake.
I'm not complaining, just reporting. My guess is that if I didn't have the offline maps I would never have noticed a change from 3.53 to 3.54. Cheers.
Quote from Guido on 02.03.2024, 08:48Hmmm... the only possible explanation that I can think of, is that it was using the online routing engine when you were creating the routes and when you recalculated them before updating your offline maps.
Why would it use the online routing engine? Because possibly your offline maps got corrupted, causing offline calculation to fail and it falling back to online calculation.
Why was the route different today then yesterday? Because online routing engine takes long term closures and traffic incidents into account (this is a bug in the HERE framework... even though I configered the online engine to ingore long term closures and incidents, it still takes them into account... I'm in contact with HERE about this).
Why did it calculate different with the updated offline map. Because with the map update, the offline engine started working again, and thus was not taking closures and incidents into account anymore.
This is just a theory... I'll know more when I've figured out the "ignoring closures and incidents with online routing engine" with HERE.
Hmmm... the only possible explanation that I can think of, is that it was using the online routing engine when you were creating the routes and when you recalculated them before updating your offline maps.
Why would it use the online routing engine? Because possibly your offline maps got corrupted, causing offline calculation to fail and it falling back to online calculation.
Why was the route different today then yesterday? Because online routing engine takes long term closures and traffic incidents into account (this is a bug in the HERE framework... even though I configered the online engine to ingore long term closures and incidents, it still takes them into account... I'm in contact with HERE about this).
Why did it calculate different with the updated offline map. Because with the map update, the offline engine started working again, and thus was not taking closures and incidents into account anymore.
This is just a theory... I'll know more when I've figured out the "ignoring closures and incidents with online routing engine" with HERE.
Quote from Barkinmadd on 02.03.2024, 15:38Thanks Guido. I’ll give it another try by first removing my offline map to see if the behaviour changes back. Good theory.
Add - I deleted some of the offline maps (Ohio, Pennsylvania) where the odd behaviour occurred. The route was unaffected (it still looks fine). I don't know what was happening. I guess we'll see if anyone else detects sudden changes to routes they were previously working on...
Thanks Guido. I’ll give it another try by first removing my offline map to see if the behaviour changes back. Good theory.
Add - I deleted some of the offline maps (Ohio, Pennsylvania) where the odd behaviour occurred. The route was unaffected (it still looks fine). I don't know what was happening. I guess we'll see if anyone else detects sudden changes to routes they were previously working on...
