Dane, I tried ST_Union against the non-overlapping version of shoreline_300 which was loaded into table lo2 unaltered. I ended up with a topology exception.
gis=# SELECT ST_AsText(ST_Multi(ST_Union(f.geom))) AS geom FROM lo2 AS f; NOTICE: TopologyException: found non-noded intersection between -7.71444e+06 -7.46419e+06, -7.71472e+06 -7.47177e+06 and -7.71446e+06 -7.47097e+06, -7.71501e+06 -7.47092e+06 -7.71469e+06 -7.47095e+06 geom ------ (1 row) Perhaps working with the source to shoreline_300 would be better. Is the source available? [Judging from the openstreetmap wiki, the data comes from an NGA website. However, one of the files, shapefile13e is missing. Not sure what to do about this.] Brad From: Dane Springmeyer [mailto:[email protected]] Sent: Monday, September 20, 2010 1:51 PM To: Simpson Brad-C-Lockheed Cc: mapnik-users Subject: Re: [Mapnik-users] Projection problems http://postgis.refractions.net/documentation/manual-1.5/ST_Collect.html http://postgis.refractions.net/documentation/manual-1.5/ST_Union.html On Sep 20, 2010, at 8:51 PM, Simpson Brad-C-Lockheed wrote: Dane, I should have read all of your Friday's email. I'm not trying to hide lines between tiles. Actually, I'm interested in collecting the geometries so that no tiles exists, in effect generating coastlines. Is there a PostGIS operation which can do this? Alternatively, the source of the data used to generate shoreline_300 may be a better place to start. Is the source data for shoreline_300 available on the web? Brad From: Dane Springmeyer [mailto:[email protected]] Sent: Monday, September 20, 2010 12:01 PM To: Simpson Brad-C-Lockheed Cc: mapnik-users Subject: Re: [Mapnik-users] Projection problems Brad, Try setting gamma=".6" in your PolygonSymbolizer as per: http://trac.mapnik.org/wiki/PolygonSymbolizer#DefaultfillwithGammacorrection Dane On Sep 20, 2010, at 7:23 PM, Simpson Brad-C-Lockheed wrote: Dane, The no_overlap* files are what the name implies - no overlapping of the tiles. However, I still can't plot shorelines without getting a crosshatch pattern because of the non-overlapping tiles. Is there an un-tiled version? Brad From: Dane Springmeyer [mailto:[email protected]] Sent: Thursday, September 16, 2010 6:18 PM To: Simpson Brad-C-Lockheed Cc: Jon Burgess; [email protected]<mailto:[email protected]>; Christopher Schmidt; Robert Coup; mapnik-users Subject: Re: [Mapnik-users] Projection problems On Sep 17, 2010, at 1:45 AM, Simpson Brad-C-Lockheed wrote: I'm hoping to solicit some advice regarding shoreline_300. We'd like to use it since it is an improvement over the data currently employed. We'd like to plot it with a matching coastline where we can adjust its style (color, width, etc.). If I use the LineSymbolizer (along with the PolygonSymbolizer to fill in land areas) on shoreline_300, I get in addition to a coastline, an undesired cross-hatch pattern. This is because the land polygons have been tiled. Yes. Well, because they have been tiled, with an overlap. There are a set of versions that do not have overlap that Jon produced as a test recently that might be helpful: http://tile.openstreetmap.org/no_overlap_processed_p.tar.bz2 http://tile.openstreetmap.org/no_overlap_shoreline_300.tar.bz2 (note that shoreline_300 is a simplified version of processed_p) Is there an un-tiled version of shoreline_300 available? To get an untiled version I would import the shapefile into postgis and collect the geometries (but beware this will be a very expensive operation). Hopefully you can make due with tiled files without overlap. To get rid of the slight lines that will appear in the non-overlapping version try setting gamma=".7" in your PolygonSymbolizer. Is there a matching coastline_300 file? If not, what do you recommend for getting around this problem? Any advice would be appreciated. Thanks, Brad -----Original Message----- From: Jon Burgess [mailto:[email protected]] Sent: Wednesday, September 15, 2010 11:51 AM To: Simpson Brad-C-Lockheed Cc: Christopher Schmidt; Dane Springmeyer; mapnik-users; Robert Coup; [email protected]<mailto:[email protected]> Subject: RE: [Mapnik-users] Projection problems On Wed, 2010-09-15 at 09:34 -0600, Simpson Brad-C-Lockheed wrote: Chris, Thanks for responding. I understand the spatialreference map is coarse and low resolution. However, my position is all coordinates in the general area are shifted (although I have only looked in the English Channel area and the Greek Islands so far). And the vmap0 data matches our database. I've only given one example that is easy to confirm. I do not believe this is a case of the data being too coarse. I would like to migrate from our old database (heritage unknown) to openstreetmap, but I can't unless I can account for this discrepancy. This map comparison between OSM and Google satellite shows a good match: http://tools.geofabrik.de/mc/?mt0=mapnik&mt1=googlesat&lon=-2.22122&lat=49.701&zoom=16 This comparison with Yahoo satellite data also shows a good match: http://sautter.com/map/?zoom=13&lat=49.70034&lon=-2.22001&layers=00000BTFFFFFFF It looks like the footpath shown in OSM has been uploaded as a GPX trace which seems to add further confirmation that the island is probably shown in the correct position in OSM. I guess you can look for more 3rd party data to confirm the location, or failing that, you'll need to go there yourself :) Jon _______________________________________________ Mapnik-users mailing list [email protected]<mailto:[email protected]> https://lists.berlios.de/mailman/listinfo/mapnik-users
_______________________________________________ Mapnik-users mailing list [email protected] https://lists.berlios.de/mailman/listinfo/mapnik-users
