[OSM-talk] Coastline?

Michal Migurski mike at stamen.com
Sat Sep 6 17:54:32 BST 2008


Thanks,

Is this the main site?
	http://hypercube.telascience.org/~kleptog/

It's not immediately obvious from the wiki. Could those shapefiles be  
available via the planet directory?

-mike.


On Sep 5, 2008, at 3:55 AM, Martijn van Oosterhout wrote:

> The coastline checker as a side effect produces the shapefiles for  
> mapnik.
>
> The reason is that the coastline in the database isn't a geometry, but
> a collection of ways. The checker joins all the segments into several
> million node linestrings and then chops it into peices again and makes
> shapes mapnik can render.
>
> It's possible the main site hasn't updated the shape in a while, so if
> you've added stuff recently and it isn't appearing you might need to
> prod someone.
>
> Have a nice day,
>
> On Fri, Sep 5, 2008 at 3:12 AM, Michal Migurski <mike at stamen.com>  
> wrote:
>> Hi,
>>
>> I'm trying to understand how OSM generates coastlines in Mapnik, to
>> render my own tiles.
>>
>> I'm reading this:
>>       http://wiki.openstreetmap.org/index.php/Coastline#Main_Mapnik_Layer
>>
>> It sounds as though there should be some presence of  
>> natural=coastline
>> in the osm.xml stylesheet:
>>       http://trac.openstreetmap.org/browser/applications/rendering/mapnik/osm.xml
>>
>> Instead, I see shapefiles. Are those derived from data in pgsql, or
>> are they obtained from elsewhere? Why are they shapefiles instead of
>> geometries from the database?
>>
>> -mike.
>>
>> ----------------------------------------------------------------
>> michal migurski- mike at stamen.com
>>                 415.558.1610
>>
>>
>>
>>
>> _______________________________________________
>> talk mailing list
>> talk at openstreetmap.org
>> http://lists.openstreetmap.org/listinfo/talk
>>
>
>
>
> -- 
> Martijn van Oosterhout <kleptog at gmail.com> http://svana.org/kleptog/
>

----------------------------------------------------------------
michal migurski- mike at stamen.com
                  415.558.1610







More information about the talk mailing list