[OSM-talk] Enabling communities to use OSM as a planning tool

[Dave Stubbs]

On Wed, Jun 4, 2008 at 9:26 PM, Jon Burgess <jburgess777 at googlemail.com> wrote:
> On Wed, 2008-06-04 at 11:09 -0400, simon at mungewell.org wrote:
>> At present they can achieve the same effect using off line file and
>> merging layers before rendering.
>>
>> To make the process more smooth maybe some (or all) of the following
>> can
>> be implemented:
>>
>> 1) Enable the render to use multiple '.osm' files for rendering,
>> automatically merging or laying onto the output.
>
> The osm2pgsql code used for the mapnik layer can already load multiple
> osm files into a single render database (see the --append option).
>
> Alternatively the Mapnik osm.xml file can easily pull in data from many
> different sources when rendering. The existing style already uses
> multiple database connections and shapefiles to build up the main map.
>
> If the data was in the DB then it would be a bit tricky to exclude data
> from the rendering. We would need to have prior knowledge of what tags
> to exclude from rendering.  I personally think it would make more sense
> to keep the data out of the main OSM DB.

Or keep the tag names very different. Essentially namespace them. ie:
highway(planning)=cycleway
highway(historic)=track  -- the historic space would expect the epoch
style tag mentioned before to also be on the object

we wouldn't have any trouble not rendering these, but they would be
easy to convert to standard tag names if status changed or your
renderer wished to treat them as the same/override the existing
feature.

Editing problems remain of course.

Dave