logo separator

[mkgmap-dev] Ferry line routing issue

From Vuki szun at informatik.hu on Mon Jun 5 11:06:37 BST 2023

Hi Nick,

that did not help.

Br,
Mihaly

On 2023. 06. 01. 8:53, osm wrote:
> Hi
>
> Not a real answer but perhaps do a quick check using another routable 
> type (0x16) for ferries (0x1a or 0x1b) and see what happens.
>
> r
>
> Nick
>
> On 31/05/2023 16:59, 7770 wrote:
>> Hi.
>> I can experience the same type of issues (or at least very similar) 
>> on my
>> GPSMAP 66st (handheld outdoor unit) when trying to take a ferry from the
>> Scandinavian peninsula to Europe.
>> I have seen some of the jumps (straight lines occasionally).
>>
>> Here is some theory on the issue.
>>
>> Examples for such issues are Karlskrona (south of sweden) to Gdynia 
>> (north of
>> Poland), or Stockholm to Helsinki. In those cases two the following can
>> happen.
>>
>> When calculating for time, the unit may consider to drive around 
>> instead of
>> taking the ferry, the ferry entrance would just be a stop along the 
>> route but
>> the device would not consider taking the ferry, because it is 
>> presumed slower
>> than driving many 100 kilometers extra or it simply times out saying 
>> there is
>> no route possible to calculate.
>>
>> If i instead select calculating for shorter distance, the ferry is often
>> considered and used by the route.
>>
>> Anyhow, in the style file for lines, there are two entries for ferry 
>> lines:
>> route=ferry & (motorcar=no | motor_vehicle=no) [0x1b road_class=0 
>> road_speed=0
>> resolution 23]
>> route=ferry [0x1a road_class=3 road_speed=0 resolution 19]
>>
>> If the ferry routes are expected to take cars the road_class will be 
>> 3 and and
>> the road_speed 0, this road speed is so slow that the device will 
>> consider to
>> use a faster way instead of taking the ferry.
>>
>> So, what i do to make it a greater chance that a ferry will be used when
>> routing for time, is that i have raised the road_speed to 1 for those 
>> two
>> lines, since it would also help for passenger ferries. I was tempted 
>> to raise
>> it to 2 but many ferries are too slow for road_speed 2.
>> This has somewhat improved routing scenarios in my case. It's not 100 %
>> perfect but much better than before.
>>
>> I cant say it is exactly the problem that you see, but some idea.
>>
>>
>> Regards
>> Karl
>>
>>
>>
>> On onsdag 31 maj 2023 15:58:00 CEST Vukovics Mihály wrote:
>>> Hello Community,
>>>
>>> I am struggling with an issue. I am generating Garmin map from OSM o5m
>>> sources(geofabrik) using the built-in style.
>>> Nor the BaseCamp or my Zümo 396LM is unable to calculate route via 
>>> ferry
>>> lines. The behaviour is very strange, here is an example:
>>>
>>> Route from Lille to Dover, all avoidances are disabled: unable to
>>> calculate route, just a straigth line between the citie.
>>> If I am forcing the route inside the service road area of Calais ferry
>>> terminal, the route is calculated correctly.
>>> Even funnier, that when I shape the route directly to the ferry line,
>>> Basecamp swims me on the ferry instead of routing through the port.
>>>
>>> This behaviour is occuring with other ferry lines (for example Aberdeen
>>> - Lerwick).
>>>
>>> I have checked on openstreetmap and the roads are connected and 
>>> routable
>>> in the style file.
>>>
>>> Here are the example screenshots about the issue.
>>>
>>> https://picallow.com/mkgmap-ferry-routing/?usp_success=2&post_id=116783&form 
>>>
>>> _id=27
>>>
>>> Any ideas or fix?
>>>
>>> Br,
>>> Mihaly
>>>
>>> _______________________________________________
>>> mkgmap-dev mailing list
>>> mkgmap-dev at lists.mkgmap.org.uk
>>> https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
>>
>>
>>
>> _______________________________________________
>> mkgmap-dev mailing list
>> mkgmap-dev at lists.mkgmap.org.uk
>> https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
> _______________________________________________
> mkgmap-dev mailing list
> mkgmap-dev at lists.mkgmap.org.uk
> https://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev


More information about the mkgmap-dev mailing list