frederik at remote.org
Tue May 6 22:50:38 BST 2008
> Namely, that the inner hulls (rings) of a multipolygon must have the
> inverse direction of the outer hull (ring), or else some renderers
> will bork up the result.
That's a problem with renderers and not a problem with the model.
Whether an area is drawn clockwise or counterclockwise *should* not
make a difference. It is true that you might achive better results *at
the moment* by drawing your areas in a certain direction but it is
certainly not necessary from a data model perspective.
(Unless you want to start discussing that a way loop drawn around a
forest does not exactly specify whether the little bit inside is the
forest or the whole rest of the sphere of the Earth is the forest.
That, indeed, is left open by our way of dealing with things.)
Frederik Ramm ## eMail frederik at remote.org ## N49°00'09" E008°23'33"
More information about the dev