There is a problem calculating a longer route (more than 100 km), “Motorway” disabled and “Small Local Road” enabled. The calculated route is very long and strange. With “Small Local Road” disabled, the route is ok.
I use the first setting when I planning a sightseeing trip.
Some examples with default car settings only “Motorway” disabled. Routing mode “Fastest”
1)
Departure: Denmark / Horsens/ town center. Destination: Denmark / Aabenraa / town center.
“Small Local Road” enabled . 190 km 2:29 h
“Small Local Road” disabled. 106 km 1:37 h
2)
Departure: Germany / Bremen/ town center. Destination: Germany / Kiel/ town center.
“Small Local Road” enabled . 446 km 6:36 h
“Small Local Road” disabled. 205 km 3:13 h
Only 2 examples, but it is always the same problem when the correct route is longer than 100 km.
This is tested on Galaxy S4 and MF version 2.1.97 an newest maps, but the problem also occurs in the PC version 15.0.2.
Let me guess: Again it might be, that a lot of roads used with "small local road enabled" are mapped as highway=unclassified in OpenStreetMap-database. For me that is a bug in MapfactorNavigatorFree as said several times here in forum, because unclassified is the lowest class of settlement connecting roads in OSM. I'll check it next time and will report. At the moment I'm busy jobwise.
I tested myself your second route Bremen-Kiel and got these results: Motorway disabled and small local enabled got 203 km using the ferry near Glückstadt. Motorway disabled and small local disabled got 209 km also using that ferry. Even disabling ferry got just 216 km driving through Hamburg. Last test only allowing major, secondary and local connecting got about 270 km arround south of Hamburg. So my guess was wrong and I have no idea what's wrong with your settings. Can you post a screenshot of your long route and your settings (or xml-file)? So one could analyze that.
I have uninstalled the MF Navigation Free and reinstall it again to be sure to have the reel default car settings. After I have reinstalled it, I only downloaded the German North map, and disabled “Motorway” and tried to shift between enabled and disabled “Small Local Road”, and I got same results as mentioned before, so I can’t see what should be wrong with my car settings. I didn’t restore any backup.
I also tried the Danish route on another Phone with a Danish map downloaded in marts and it found the correct route with both settings. Then I updated the Danish map and the Earth map and then I got the problem.
In my profile I always have "small local roads" enabled. However I use 30/25/50 as parameters instead of the default 60/40/50.
This to prevent sometimes "strange" routing and of course it is more according "real life" settings.
I also encountered the strange routing a couple of times in Brittany and North/Mid-England. When I reduced the speeds it was much better, also when calculating an ETA.
Yes I agree, there must be a problem (MapFactor bug) in the maps, not only the German map but also the Danish map. As told before I can find a lot of this type of problem when calculating routes I Denmark.
I hope that the good MapFactor people will come into this discussion, telling whether I'm right or wrong
I tested again but now I used a pretty customized profile I made for my 125 ccm scooter excluding motorways and special speed allover. Enabling small locals I also got this route with 446 km or so. But I don't think its a failure in map data but in algorithm of route calculation. A map bug would also work with other settings. Maybe nobody found out this because it's unusual to make trips, where small local roads should be used for driving exept the first and last miles from start or to target, if there is no other way.
If Mapfactor would aggree to my - and others - proposal to put unclassified mapped roads into a higher class, because they are the lowest kind of settlement connecting roads, there wouldn't be no need to activate small locals for trips.
I still think the problem has someting to do with the maps, se my comment Jun 17. May be it has something to do with the “urban script” because, the first time I recognized it was when I have downloaded the first “urban script” Danish map
@hurdygurdyman for witch car settings (with “Small local roads” enabled) do you not see the Bremen to Kiel problem, I would like to test it, to see if it also is ok when I am testing it. Tank you in advance.
I did some tests as well with the Dutch map while excluding motorways. I really get the strangest routes when calculating routes longer than 200 km.
But also sometimes they are (or at least seem) correct.
I guess hurdygurdyman is right in that it is a bug in the algorithm.
For this thread more or less off topic but also algorithm related:
And as a side note for which I was doubting to create a topic: Last week I was on holiday and I was in a couple of big cities. I also see that Navigator has problems there. It always calculates the seemingly fastest way, but it doesn't. If I use apps that use "speed penalties" (traffic lights, speed (reducing) bumps, crossroads, etc.) I do get a different route that is indeed better.
Navigator only checks the speed for a road segment without looking at anything else. It caclulates travel times (in cities) that are in general half of what they should be and which are calculated (correctyl!) by some of the others. Note that Navmii and Maps.me don't use speed penalties either and do get the same way too short travel times as well. OsmAnd does this by far the best.
Of course on long trips outside cities this doesn't matter: a 3-4 hours drive with a calculated 10 or real 20 minutes in a city (at start and end), is not much of a difference.
Don't forget that 2.1.97 was released on 13 April. The bug can be from both parts: map & app.
Unless the April maps were also released 3rd week of April. In that case it was after the app release unless you updated/upgraded very late.But that can also be valid fro maps, which might be of March when you did your previous tests.
Can you remember map dates and app upgrade date?
The June maps will arrive this week. We'll see wether the bug is gone, or wther it continues.
The May map was released on May 28 (data version 44-201605280). The previous map (that works ok) was likely released on April 2 (data version 44-201604020).. I always download updates as soon as they are available. Let's wait for the next map update.
yet routes are correct if I use the April map and they are not if I use the May or June map. The 'reason' is in the combination of maps and the way how the app manages the modified car settings. Let's be patient and wait for a solution.