MapFactor Forum

Uli

By missing I mean is not colored blue, it's just a criss-cross of ship routes.

What is worse, the whole western shore is missing together with a lot of details. Don't know if it is an OSM-map or a rendering problem.

Uli

chattiewoman

Hi Uli,
Ob es ein OSM-Fehler ist, kannst du unter osm.org selber nachsehen. Einfach nach dem Namen suchen oder manuell heranzoomen.

SunCobalt

Looks okay to me
http://ra.osmsurround.org/analyzeMap?relationId=168893

retsmoor

Looks also ok to me ...

Uli

Well since last week's map update here is how it looks on all my 3 navigator versions:

Windows and WINCE 12.3.0, Android 11.0.37

retsmoor

... my versions :
- WINDOWS & WINCE 12.3.0
- ANDROID 1.0.35
- Germany South OSM July 19, 2013

retsmoor

UPDATE :
Germany South OSM Update version 201309240 via "Early maps access" ...
I can confirm this issue ...

Oldie

Same hier on the Android Version with September maps.

mdx

I can confirm that this relation was rejected (among others from this area)
getWaterShp: processing square 'gaaaacadaadbc'
Bad water relation/polygon 20470 -
Bad water relation/polygon 168893 - Ammersee
Bad water relation/polygon 240935 -
Bad water relation/polygon 1429361 -
Bad water relation/polygon 1429362 -
but I am not sure now, what was exactly the problem there
more later
Martin

mdx

Well, for 90% the problem is in "invalid geometry", i.e. you have polygon with holes but the holes are connected and do not define "compact hole". Following test fails:
def testBugAmmersee( self ):
    # relation 168893
    poly = nodeList2Polygon( [ [(0,0,1), (100,0,2), (100,100,3), (0,100,4), (0,0,1)],
      [(10,10,5),(50,10,6),(10,50,7),(10,10,5)],
      [(50,10,6),(10,50,7),(50,50,8),(50,10,6)] ] )
    self.assert_( poly.is_valid )
and it is case of Ammersee too (there are three connected holes/islands). It is the south part of the lake:
http://www.openstreetmap.org/browse/way/207090745
regards
   Martin

SunCobalt

we call it touching inner rings. This is an error in standard GIS definition but not in OSM. Some in OSM wants to be more in line with the standard GIS definition and we have a map showing touching inner
http://tools.geofabrik.de/osmi/?view=multipolygon&lon=10.98340&lat=51.85885&zoom=7&overlays=invalid_geometry_hull,duplicate_ways,intersections,intersection_lines,ring_not_closed_hull,ring_not_closed,unconnected_end_nodes,touching_inner_rings_hull,touching_inner_rings,role_mismatch_hull,role_mismatch,duplicate_tags_hull,duplicate_tags,multipolygons_type_is_boundary,type_is_boundary,ways,role_markers,way_end_nodes,way_nodes

But as of now, touching inner rings in a multipolygon are not an error and should not prevent a map from displaying the object

see OSM Wiki
http://wiki.openstreetmap.org/wiki/Multipolygon#Touching_inner_rings

mdx

OK, thanks for info. We will try to fix that, i.e. merge the holes during processing.

mdx

Ammersee should be fixed now (i.e. cases with polygon containing several touching holes). OTOH the remaining cases, like polygon with touching hole (http://www.openstreetmap.org/browse/relation/240935) is not fixed yet, and it will require thorough review/analysis.
thanks
Martin

SunCobalt

http://www.openstreetmap.org/browse/relation/240935
^^ is not a valid Multipolygon, even in OSM ;) The inner rings are no allowed to overlap or touch the outer ring. And the area tag (here graveyard should be attribute of the relation, not at the outer ring.
I would not waste time with obviously wrong objects. I have fixed it

MapFactor Forum

Howdy, Stranger!

Categories

In this Discussion