"Robert (Jamie) Munro" <[EMAIL PROTECTED]> writes: > Dirk-Lüder Kreie wrote: >> Maarten Deen schrieb: >>> I just commited a change for oceantiles.dat where I marked all tiles >>> as mixed if there exists a coastline node in it for tiles of latitude >>> 60 and higher. >>> The coastline is not nearly complete, but this will give a nice start >>> for people who care about it. ;) >> >> you do know that this contradicts the schema of how the oceantiles >> hinting works? >> >> http://wiki.openstreetmap.org/index.php/Tiles%40home/How_Oceantiles_work >> >> For example a group of islands will now render inverted (again!) in >> certain cases. > > Islands are supposed to render based on the principle that water is on > the right. As I understood things to work, wherever there is any kind of > coastline in a z12 tile, oceantiles.dat is not even consulted.
Without looking at close-areas.pl this doesn't seem to be the case. Islands on a tile where the coastline does not cross tile borders are rendered inverted. We might call that a bug in close-areas. Matthias _______________________________________________ Tilesathome mailing list [email protected] http://lists.openstreetmap.org/listinfo/tilesathome
