From: Michael B. <mic...@gm...> - 2009-05-13 13:57:16
|
> So I think that the only solution is to generate, at every rendering, a > buffer > the size of the image that stores for each pixel the feature id of the last > rendered feature on it. > Yikes !!! That's starting to sound really unpleasant for you. Perhaps a more low-tech approach would be better :) I think you are working with point features - yes ? How about just converting the mouse click to world coordinates and looking for features that are within a threshold distance (in world units) to that location. You could base the threshold on some small fraction of the currently display map extent. If you catch more than one feature just ask the user to zoom in. Hope this helps Michael |