[OSM-talk] Topology (was: OSM the mediocre alternative)

[Christopher Schmidt]

On Mon, Apr 23, 2007 at 12:01:19AM +0200, Frederik Ramm wrote:
> Hi,
> 
> Johnny Doe <uucp1 at yahoo.com> wrote:
> > please read about the GRASS vector data model before
> > claiming a fundamental incompatibility between Simple
> > Features and a topological network:
> > http://grass.itc.it/grass62/manuals/html62_user/vectorintro.html
> 
> I never claimed a fundamental incompatibility! I just thought 
> Christopher wanted to say we don't need topology except for routing, and 
> answered I believe we also need topology for editing.

I believe that topology for editing is best implemented in the *clients*, 
rather than the server.

The model that I visualize has a couple things:
 * clients are topological. That is, they download data from the API,
   and that data is such that it is a topological mesh. By dragging
   points onto one another, they 'snap' into place: snapping being done
   on the client side. (Note that although OSM is topological, it
   doesn't enforce this topology in any way... even JOSMlint doesn't
   seem to complain.)
 * Data with non-trivial topological information -- 'this junction
   doesn't allow left turns' -- is an additional node at a junction, or 
   border, or whatever you need.
   
Perhaps the answer is 'it is too hard for clients to create topology
from simple features' -- but why is that? A single point which is
covered by two gemetries is an intersection. What would make this
difficult on the clientside?   

Regards,
-- 
Christopher Schmidt
MetaCarta