[cig-commits] commit: There appears to be some kind of problem with either

Mercurial hg at geodynamics.org
Mon Nov 24 11:28:56 PST 2008


changeset:   19:3a0aa63d14c3
user:        LukeHodkinson
date:        Thu Aug 02 01:20:49 2007 +0000
files:       DrawingObjects/src/HistoricalSwarmTrajectory.c
description:
There appears to be some kind of problem with either
the parallel compositing algorithm or the way we're
setting the depth of each drawing object. When proc 1
draws something in proc 0's domain, proc 0 clobbers
it as a result of the depth comparison.  I've modified
the depth offset at which the historical swarm trajectories
get rendered such that they now appear on proc 0; the odd
thing is that I had to move them further away from the
camera for this to work: the opposite to what you'd
expect.  Anyway, seems to work now.


diff -r 53a543dad70c -r 3a0aa63d14c3 DrawingObjects/src/HistoricalSwarmTrajectory.c
--- a/DrawingObjects/src/HistoricalSwarmTrajectory.c	Fri Jul 20 06:09:06 2007 +0000
+++ b/DrawingObjects/src/HistoricalSwarmTrajectory.c	Thu Aug 02 01:20:49 2007 +0000
@@ -39,7 +39,7 @@
 *+		Patrick Sunter
 *+		Greg Watson
 *+
-** $Id: HistoricalSwarmTrajectory.c 717 2007-07-20 06:09:06Z LukeHodkinson $
+** $Id: HistoricalSwarmTrajectory.c 727 2007-08-02 01:20:49Z LukeHodkinson $
 ** 
 **~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~*/
 
@@ -256,7 +256,7 @@ void _lucHistoricalSwarmTrajectory_Build
 	Particle_Index                 lParticle_I;
 	StandardParticle*              particle;
 	lucHistoricalSwarmTrajectory_ParticleExt* particleExt;
-	float                          offset              = 0.001;
+	float                          offset              = -1.0;
 	double*                        coord;
 	Dimension_Index                dim                 = context->dim;
 
@@ -282,7 +282,7 @@ void _lucHistoricalSwarmTrajectory_Build
 
 			if (dim == 2)
 				glVertex3f( (float)coord[0], (float)coord[1], offset );
-			else 
+			else
 				glVertex3dv( coord );
 				
 			/* Stop the loop when we have arrived a the current timestep */



More information about the CIG-COMMITS mailing list