[mkgmap-dev] mkgmap:set_unconnected_type differentiate between connected on both sides or on one side only

Thu Nov 14 13:19:05 GMT 2019

Hi Felix,

no idea why you would modify to code instead of using option --min-size-polygon=14, so I agree this is not important.

I was surprised to find that there is no documentation about mkgmap:set_unconnected_type. I found a lot of confusing posts about the expected behaviour, so it's very difficult to guess what it really should do. With the additional tag mkgmap:set_semi_connected_type it get's even more complicated.
Since you wrote that the patch worked well for you I think we should just document what it does.

The patch was written before option --add-boundary-nodes-at-admin-boundaries was implemented.
A road with a node on a tile boundary is not removed when mkgmap:set_unconnected_type=none (or mkgmap:set_semi_connected_type=none) is found.
I think we can ignore nodes on country borders here (like we do with the dead-end-check.

My understanding so far, please improve:
The algorithm first finds out how often a road (one OSM way) is connected to other roads. If there are zero connections and mkgmap:set_unconnected_type=none  is set, the way is removed (not written to the map). If there are other (overlaying) lines for the same OSM way and they also have mkgmap:set_unconnected_type=none they are also removed.
Similar things happen with tag mkgmap:set_semi_connected_type=none for roads with exactly one connection to another road.

There is also an option to specify a different type instead of none, e.g. mkgmap:set_unconnected_type=0x10804. This type must be a non-routable line type. The effect is that the line is not added as a road with the original type, instead it is added as normal line with the fiven type when the road is not connected to other roads. Similar with mkgmap:set_semi_connected_type=0x10804.
The current patch doesn't care about overlay lines when a replacement type is given for the road, means, they are neither removed nor changed. Looks wrong to me.
If I got that right you only use tag values "none" and nobody else uses this feature, so I'd prefer to remove the type change. In that case it would be better to also rename the tags, e.g. mkgmap:remove-unconnected=true and mkgmap:remove_semi_connected=true

Any thoughts?

Gerd

________________________________________
Von: mkgmap-dev <mkgmap-dev-bounces at lists.mkgmap.org.uk> im Auftrag von Felix Hartmann <extremecarver at gmail.com>
Gesendet: Mittwoch, 13. November 2019 19:29
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] mkgmap:set_unconnected_type differentiate between connected on both sides or on one side only

Hi Gerd,
Nope that's the only one. Until yesterday I think it worked correctly. I never really check again in the last 1 year however.  And I do not use --housenumber option so far. It's very useful to filter out private ways which are not marked access=private - I mainly use it on highway=footway/path/track and so on (which should not have housenumbers anyhow).

(well and what I hardly call a patch - + minSizePolygon = props.getProperty("min-size-polygon", 14);
value 14 instead of 8. I think 8 is too small in src/uk/me/parabola/mkgmap/build/MapBuilder.java

On Wed, 13 Nov 2019 at 19:00, Gerd Petermann <gpetermann_muenchen at hotmail.com<mailto:gpetermann_muenchen at hotmail.com>> wrote:
Hi Felix,

I already wondered where this code has gone. :o
Hard to say if it is still useful. I probably would not want to use it in combination with option --housenumbers.
I am able to modify the patch. Are there any other patches from me or others that you use?

Gerd

________________________________________
Von: mkgmap-dev <mkgmap-dev-bounces at lists.mkgmap.org.uk<mailto:mkgmap-dev-bounces at lists.mkgmap.org.uk>> im Auftrag von Felix Hartmann <extremecarver at gmail.com<mailto:extremecarver at gmail.com>>
Gesendet: Mittwoch, 13. November 2019 17:12
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] mkgmap:set_unconnected_type differentiate between connected on both sides or on one side only

Could semi_con-v3.patch so the semi_connected tag be merged to trunk? It was compatible until now and is of really good use. Unfortunately - it conflicts with todays mkgmap update.
Sorry I forgot to answer back in 2017 confirming that it worked splendidly.

Felix

On Thu, 21 Sep 2017 at 15:02, Felix Hartmann <extremecarver at gmail.com<mailto:extremecarver at gmail.com><mailto:extremecarver at gmail.com<mailto:extremecarver at gmail.com>>> wrote:
Thanks for v3 - on a quick tryrout it works well now. I have not found time (and won't until Tuesday next week) to give it a full check.
(and yes - I'm right now only using "none").

On 21 September 2017 at 11:56, Gerd Petermann <GPetermann_muenchen at hotmail.com<mailto:GPetermann_muenchen at hotmail.com><mailto:GPetermann_muenchen at hotmail.com<mailto:GPetermann_muenchen at hotmail.com>>> wrote:
Hi Felix,

sorry, did not search for a solution for my example, what I wanted to point out is that the algo may produce unexpected results whenever the style adds multiple lines for one way with conflicting mkgmap:set_semi_connected_type values.

In your style you can probably only use value "none". I wonder if anybody uses the variant with a value that gives a different type.

Gerd

________________________________________
Von: mkgmap-dev <mkgmap-dev-bounces at lists.mkgmap.org.uk<mailto:mkgmap-dev-bounces at lists.mkgmap.org.uk><mailto:mkgmap-dev-bounces at lists.mkgmap.org.uk<mailto:mkgmap-dev-bounces at lists.mkgmap.org.uk>>> im Auftrag von Felix Hartmann <extremecarver at gmail.com<mailto:extremecarver at gmail.com><mailto:extremecarver at gmail.com<mailto:extremecarver at gmail.com>>>
Gesendet: Donnerstag, 21. September 2017 11:16
An: Development list for mkgmap
Betreff: Re: [mkgmap-dev] mkgmap:set_unconnected_type differentiate between connected on both sides or on one side only

As for your example - yes I guess only changing first occurence to 0x* - further occurences to none makes most sense. In general I think such a rule should not be used.

So good practice would be either:
highway=service & service=driveway {set
mkgmap:set_semi_connected_type=none}
highway=service [0x07 road_class=0 road_speed=2 resolution 22 continue]
highway=service & oneway=yes [0x10106 resolution 24]

or
highway=service & service=driveway {set
mkgmap:set_semi_connected_type=0x10806} [0x07 road_class=0 road_speed=2 resolution 22 continue]
highway=service [0x07 road_class=0 road_speed=2 resolution 22 continue]
highway=service & oneway=yes [0x10106 resolution 24]

or
highway=service & service=driveway {set
mkgmap:set_semi_connected_type=0x10806}
highway=service [0x07 road_class=0 road_speed=2 resolution 22 continue]
highway=service & service=driveway {set
mkgmap:set_semi_connected_type=0x10806}
highway=service & oneway=yes [0x10106 resolution 24]

but not your example and also not:
highway=service & service=driveway {set
mkgmap:set_semi_connected_type=0x10806} [0x07 road_class=0 road_speed=2 resolution 22 continue with_actions]
highway=service [0x07 road_class=0 road_speed=2 resolution 22 continue]
highway=service & oneway=yes [0x10106 resolution 24]

On 21 September 2017 at 11:10, Felix Hartmann <extremecarver at gmail.com<mailto:extremecarver at gmail.com><mailto:extremecarver at gmail.com<mailto:extremecarver at gmail.com>><mailto:extremecarver at gmail.com<mailto:extremecarver at gmail.com><mailto:extremecarver at gmail.com<mailto:extremecarver at gmail.com>>>> wrote:
Somthing seems to be wrong with the patch:

java.lang.NullPointerException
        at uk.me.parabola.mkgmap.osmstyle.StyledConverter.findUnconnectedRoads(StyledConverter.java:1970)
        at uk.me.parabola.mkgmap.osmstyle.StyledConverter.end(StyledConverter.java:605)
        at uk.me.parabola.mkgmap.reader.osm.ElementSaver.convert(ElementSaver.java:243)
        at uk.me.parabola.mkgmap.reader.osm.OsmMapDataSource.load(OsmMapDataSource.java:157)
        at uk.me.parabola.mkgmap.main.MapMaker.loadFromFile(MapMaker.java:154)
        at uk.me.parabola.mkgmap.main.MapMaker.makeMap(MapMaker.java:52)
        at uk.me.parabola.mkgmap.main.Main$1.call(Main.java:263)
        at uk.me.parabola.mkgmap.main.Main$1.call(Main.java:259)
        at java.util.concurrent.FutureTask.run(Unknown Source)
        at java.util.concurrent.ThreadPoolExecutor.runWorker(Unknown Source)
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(Unknown Source)
        at java.lang.Thread.run(Unknown Source)
Could Not Find C:\OpenMTBMap\maps\ovm_6431*.img

On 19 September 2017 at 15:53, Gerd Petermann <gpetermann_muenchen at hotmail.com<mailto:gpetermann_muenchen at hotmail.com><mailto:gpetermann_muenchen at hotmail.com<mailto:gpetermann_muenchen at hotmail.com>><mailto:gpetermann_muenchen at hotmail.com<mailto:gpetermann_muenchen at hotmail.com><mailto:gpetermann_muenchen at hotmail.com<mailto:gpetermann_muenchen at hotmail.com>>>> wrote:
Attached is v2 of the patch. It implements the removal of overlay lines when
mkgmap:set_unconnected_type=none or mkgmap:set_semi_connected_type=none was
found.

I am still not sure what should be done if the tag has a value that gives
another type instead of none.
Assume your style uses
highway=service & service=driveway {set
mkgmap:set_semi_connected_type=0x10806}
highway=service [0x07 road_class=0 road_speed=2 resolution 22 continue]
highway=service & oneway=yes [0x10106 resolution 24]

What would you expect for a semi connected way?
We have 2 lines, the first is changed from 0x07 to 0x10806. It would not
make much sense to change also the 2nd from 0x10106  to 0x10806.
So, for now only the value none has an effect for the overlay line(s).

semi_con-v2.patch
<http://gis.19327.n8.nabble.com/file/t318326/semi_con-v2.patch>

Gerd

Felix Hartmann-2 wrote
> That sounds good
>
> On Sep 19, 2017 11:23 AM, "Gerd Petermann" <

> gpetermann_muenchen@

> >
> wrote:
>
>> Hi Felix,
>>
>> Felix Hartmann-2 wrote
>> > Well I would like it to apply to non routable lines too - if continue
>> > with_actions is used - basically just treat routable and non routable
>> > lines
>> > the same (the initial check should only look at routable lines though I
>> > guess).
>>
>> OK, I think I can change the code so that it stores the information
>> whether
>> or not a road
>> is connected (or "semi-connected") once for each OSM way that is at least
>> added once as a road.
>> In a further step mkgmap would check each line for the existence of the
>> mkgmap:set_unconnected_type tag and check if the corresponding OSM way is
>> connected or not.

--
Sent from: http://gis.19327.n8.nabble.com/Mkgmap-Development-f5324443.html
_______________________________________________
mkgmap-dev mailing list
mkgmap-dev at lists.mkgmap.org.uk<mailto:mkgmap-dev at lists.mkgmap.org.uk><mailto:mkgmap-dev at lists.mkgmap.org.uk<mailto:mkgmap-dev at lists.mkgmap.org.uk>><mailto:mkgmap-dev at lists.mkgmap.org.uk<mailto:mkgmap-dev at lists.mkgmap.org.uk><mailto:mkgmap-dev at lists.mkgmap.org.uk<mailto:mkgmap-dev at lists.mkgmap.org.uk>>>
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

--
Felix Hartman - Openmtbmap.org & VeloMap.org
Schusterbergweg 32/8
6020 Innsbruck
Austria - Österreich

--
Felix Hartman - Openmtbmap.org & VeloMap.org
Schusterbergweg 32/8
6020 Innsbruck
Austria - Österreich
_______________________________________________
mkgmap-dev mailing list
mkgmap-dev at lists.mkgmap.org.uk<mailto:mkgmap-dev at lists.mkgmap.org.uk><mailto:mkgmap-dev at lists.mkgmap.org.uk<mailto:mkgmap-dev at lists.mkgmap.org.uk>>
http://www.mkgmap.org.uk/mailman/listinfo/mkgmap-dev

--
Felix Hartman - Openmtbmap.org & VeloMap.org
Schusterbergweg 32/8
6020 Innsbruck
Austria - Österreich

--
Felix Hartman - Openmtbmap.org & VeloMap.org
Schusterbergweg 32/8
6020 Innsbruck
Austria - Österreich

--
Felix Hartman - Openmtbmap.org & VeloMap.org
Schusterbergweg 32/8
6020 Innsbruck
Austria - Österreich