- mkgmap-dev - The mkgmap lists

Weird MultiPolygon intersection warning
by Marko Mäkelä 02 Feb '10

02 Feb '10

Hi WanMil, can you please investigate this: 2010/02/01 09:05:31 WARNING (MultiPolygonRelation): 63240001.osm.gz: Polygon 4611686018427388326 intersects itself. 2010/02/01 09:05:32 WARNING (MultiPolygonRelation): 63240001.osm.gz: The polygon is composed of 2010/02/01 09:05:32 WARNING (MultiPolygonRelation): 63240001.osm.gz: - http://www.openstreetmap.org/browse/way/21252252 I checked http://www.openstreetmap.org/browse/relation/4370 and it looks OK to me. Three edges in the multipolygon are close to each other (two of them within the inner way 21252252) but the lines are not intersecting. Marko

2 4

Style branch
by Steve Ratcliffe 02 Feb '10

02 Feb '10

Hi I will soon merge the style branch to trunk. With the default style, you get exactly the same map with trunk and the branch and it runs at about the same speed. If however you have a more complex style there may be some differences and it may run a little slower, depending on how complex the style is. So if you have such a style, please try it out. Style should now work as documented, in that they are applied from the top of the file downward until there is a match. If you built your style on that basis then it should still work. If however you worked around the randomness that existed, you may find that the work arounds no longer work. ..Steve

2 1

Commit: r1537: Initial code for the SRT file.
by svn commit 01 Feb '10

01 Feb '10

Version 1537 was commited by steve on 2010-01-29 13:26:12 +0000 (Fri, 29 Jan 2010) Initial code for the SRT file. Not used by anything else yet.

3 3

Commit: r1546: add --show-matches to style tester
by svn commit 01 Feb '10

01 Feb '10

Version 1546 was commited by steve on 2010-02-01 17:56:28 +0000 (Mon, 01 Feb 2010) BRANCH: style add --show-matches to style tester remove wildcards from the Tags iterator. make tests pass again

1 0

Commit: r1544: Integrate continue with_actions to style branch.
by svn commit 01 Feb '10

01 Feb '10

Version 1544 was commited by steve on 2010-02-01 11:41:02 +0000 (Mon, 01 Feb 2010) BRANCH: style Integrate continue with_actions to style branch. This includes duplicating nodes/ways that are resulting in more than one element in the map. I belive that the branch is now feature comparable with the trunk. Only light testing so far.

3 5

Patch for splitter rounding
by Mark Burton 01 Feb '10

01 Feb '10

I tried to commit the attached patch that changes the splitter's area rounding to match mkgmap's new behaviour when rounding the overview map bounding box but I wasn't allowed to do it. Could someone please commit this for me. Mark

2 2

Commit: r1543: More specific warnings for multipolygons
by svn commit 31 Jan '10

31 Jan '10

Version 1543 was commited by marko on 2010-01-31 22:16:31 +0000 (Sun, 31 Jan 2010) More specific warnings for multipolygons (intersections, exchanged inner/outer roles). Self intersecting polygons are now handled as multiple polygons instead of using only the first polygon Inner polygons that are not contained by any outer polygons are dismissed now. There is a high probability that inner/outer roles are exchanged or that the inner polygon does not belong to the multipolygon. Remove polygon tags from unclosed ways because sometimes they are closed later in mkgmap. WanMil

1 0

[PATCH v1] More specific log messages in mp
by WanMil 31 Jan '10

31 Jan '10

* More specific warnings for multipolygons (intersections, exchanged inner/outer roles) * Self intersecting polygons are now handled as multiple polygons instead of using only the first polygon * Inner polygons that are not contained by any outer polygons are dismissed now. There is a high probability that inner/outer roles are exchanged or that the inner polygon does not belong to the multipolygon. * Remove polygon tags from unclosed ways because sometimes they are closed later in mkgmap. WanMil

2 2

Bug in rounding when using level 0 with resolution 23
by Felix Hartmann 31 Jan '10

31 Jan '10

When using resoluton=23 for level=0 rounding of streets and other objects can go pretty wrong. Or is this just the general rounding problem we have? Principally I think resolution23 should be enough (4.8m accuracy, vs 2.4m accuracy on resolution24) for general use. Using resolution24 I have never seen points getting rounded wrongly (so that a street seems to be broken). Routing is not affected however (no difference at all). I don't understand how one point (a junction of a way is certainly one point), gets rounded to two different places. Neither of the two following options helps/changes anything here: It is not related to the Douglac Peucker algorithm: --reduce-point-density=0 --reduce-point-density-polygon=0 (results in no visual change). It is also not related to --adjust-turn-headings

1 0

Commit: r1540: Change the signature of resolveType so that continue
by svn commit 31 Jan '10

31 Jan '10

Version 1540 was commited by steve on 2010-01-30 13:52:44 +0000 (Sat, 30 Jan 2010) BRANCH: style Change the signature of resolveType so that continue can be implemented. Continue works. But not without actions (ie it is as if without_actions is always given).

5 5