[postgis-users] snapToGrid strange behaviour

[strk at refractions.net]

Kevin, I found the bug to be in the computation of
the bounding box of SnapToGrid result.
The fix is in CVS head.

Note that if you ever stored results from SnapToGrid
back to the database you'll need to force bbox
recomputations in order to fix it. This is done
using:
	UPDATE ... SET the_geom=addBbox(dropBbox(the_geom))  ...

You can find a new regress test in CVS head to check this
(snaptogrid.sql, snaptogrid_expected).

Let me know if you still have problems.
Thanks.

--strk;


On Thu, Jan 26, 2006 at 02:25:58PM -0800, Kevin Neufeld wrote:
> Can somebody please explain why we are getting these results?
> 
> 
> test=# select snaptogrid(the_geom, 0.001) ~= the_geom from tmp1; 
> ?column?  
> ---------- 
> f 
> (1 row) 
> 
> test=# update tmp1 set the_geom = snaptogrid(the_geom, 0.001); 
> UPDATE 1 
> test=# select snaptogrid(the_geom, 0.001) ~= the_geom from tmp1; 
> ?column?  
> ---------- 
> f