[fdo-users] How does FDO handle invalid fgf geometries?

[Mateusz Loskot]

Traian Stanev wrote:
> It gives freedom of implementation to the provider writer. Internally
> the provider can use the FDO geometry stuff, but it doesn't have to.

Traian,

Yes, this is clear.

> Also, we can disagree on what is a reasonable amount of validation,
> depending on the underlying format. For example, while Oracle may not
> like polygons with self-intersections, SHP, SDF and SQLite don't give
> a shit and will accept it.

 From that angle, I agree that implementation of, let's say, 
Geometry::Validate() is provider-specific.


BTW, looks like in the lists we've archived number of interesting 
answers that belong to FDO FAQ :-)
Voting for http://trac.osgeo.org/fdo/wiki/FAQ

Best regards,
-- 
Mateusz Loskot, http://mateusz.loskot.net
Charter Member of OSGeo, http://osgeo.org