[mkgmap-dev] highway=track with no road_class and road_speed set breaks routing ## was ## "No roads near target" bug in Schwabmünchen

Sun Apr 14 08:52:53 BST 2013

Hi all,

I did some more debugging regarding the routing problem to
"Robert-Bosch-Strasse 14".  Since this works with the default style
I tried to find out why the HousenumberGenerator produces other results with
Francos style. I found no difference.

So, it really seems that we should not allow to use 
routable types for non-routable ways when creating a map with --route.

Attached is a patch that does two things:
1) Add checks to --list-styles to warn for such potential problems
2) Produce an error message if such a case is found when creating a map with
--route

routable_types_v2.patch
<http://gis.19327.n5.nabble.com/file/n5757000/routable_types_v2.patch>  

I am not sure what types the test should check. The wiki says that 
0x01-0x13 and 0x1a-0x1b are routable, so I used these.

A compiled binary based on trunk version r2570 is here:
http://files.mkgmap.org.uk/download/106/mkgmap.jar

Please let me know if this works with your style(s).

Gerd

Felix Hartmann-2 wrote
> overlays is obsolete since "continue" and "continue with_actions" exist 
> - which doesn't mean it sometime can save some lines of code for the 
> lazy - or those who use it long before "continue" existed. (though I 
> suppose those are not many, much more people simply using it, because 
> they didn't understand how and what continue does).
> 
> However if one uses "continue" and then a routable type like [0x05 
> resolution 24] - I would suppose the same bug happens, or am I wrong?
> In that case maybe obsoleting overlays because of unknown errors and 
> missing configurability should be enacted (print warning).
> 
> 
> 
> On 13.04.2013 18:08, Bernd Weigelt wrote:
>> -----BEGIN PGP SIGNED MESSAGE-----
>> Hash: SHA1
>>
>> Am 13.04.2013 17:58, schrieb Franco Bez:
>>> As the use of routable types for road2 in overlays might break the
>>> routing (route goes to tile border and then jumps on a straight
>>> line to the target, a non existant way is used for routing), you
>>> should consider printing a  warning when the 2nd type in an overlay
>>> definition is of routable type.
>>>
>>> IMHO the routing issues with housenumbers (not all housenumbers
>>> available in address search) is a completely different story.
>> My 'overlays' is now completly empty, the routing over roundabouts
>> work as expected. Housenumber-routing works, too.
>>
>> And i found some older postings from 2011 and earlier, which descript
>> a similar problem with 'overlays'
>> IMHO is 'overlays' obsolet, because most things can be done in 'lines'
>>
>> Bernd
>> -----BEGIN PGP SIGNATURE-----
>> Version: GnuPG v2.0.19 (GNU/Linux)
>> Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/
>>
>> iEUEARECAAYFAlFpgwsACgkQwCMdlf933K8SMACYs9eUkclaNQQ2hKYsSPFQRgLd
>> ZwCgntSIQztRVeX2jopmO4jK01NtFFw=
>> =mzYa
>> -----END PGP SIGNATURE-----
>> _______________________________________________
>> mkgmap-dev mailing list
>> 

> mkgmap-dev at .org

>> http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev
> 
> -- 
> keep on biking and discovering new trails
> 
> Felix
> openmtbmap.org & www.velomap.org
> 
> _______________________________________________
> mkgmap-dev mailing list

> mkgmap-dev at .org

> http://lists.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

--
View this message in context: http://gis.19327.n5.nabble.com/No-roads-near-target-bug-in-Schwabmunchen-tp5753364p5757000.html
Sent from the Mkgmap Development mailing list archive at Nabble.com.