[udig-devel] 2 issues with temporary fixes are in trunk!!

Jesse Eichar jeichar at refractions.net
Tue Oct 24 09:59:59 PDT 2006


Thanks,

I'm obviously still testing this and other functionality related to  
the rendering.  I will still be a couple of days more I expect.  I  
will test your changes.

Thanks again for the fixes and the notification of the fixes ;),

Jesse


On 24-Oct-06, at 6:12 AM, Vitali Diatchkov wrote:

>
>
> Good day/morning!
>
> 1) BasicWMSRenderer2.calculateRequestBoundingBox(...) contained wrong
> calculations: when the user adds WMS layer to the empty map, the  
> viewport
> gets bounding box of the whole layer. This leads to wrong  
> calculations of
> BBOX calculation in specified method that in its order leads to wrong
> unproportional map image returned by MapServer (wms_wrong.jpg).
> temporary quick fix is in trunk. The right map image is wms_right.jpg.
>
>
> 2) BasicFeatureRenderer.validateBounds(..) contains mistaken behavior.
> Transforming bounding box from layer CRS to WGS_84 and back  
> violates the
> accuracy. This leads to overlapping effect on the map during re- 
> rendering of
> the one feature by its bounding box when transparency is used in  
> style (see
> validatebounds_error.jpg).
> On the image (I highlighted the feature by drawing command) you can  
> easily
> see the exact bounding box of the feature that was cleared from buffer
> image, but really rendered bounding box is a bit expanded because  
> of wrong
> validateBounds(..) method calculations. And because of layer style
> transparency (filling) we see that effect of overlapping. The layer  
> CRS is
> KKJ (Finland uniform), double transformation looses the accuracy.
>
> Quick temporary fixes are in trunk in these classes. Please check  
> the issues
> to get rid of described problems.
>
> Vitali.
>
> <validatebounds_error.gif>
> <wms_error.gif>
> <wms_right.gif>
> _______________________________________________
> User-friendly Desktop Internet GIS (uDig)
> http://udig.refractions.net
> http://lists.refractions.net/mailman/listinfo/udig-devel



More information about the udig-devel mailing list