Re: [JAVA3D] Efficiency (follow up)

[Nathan Bower]

it could be garbage collection is running at those points.   do you compile your graph? what attributes do you set? Can you perform this transform somewherere farther up the scene graph - ie transform only 1 node and the subnodes automatically show the canges?   As a last resort I would download a profiler (search google) and see what exactly is using the cpu when the synchroniser "stalls".   Rgds,   N ----- Original Message ----- From: Bura Tino To: [EMAIL PROTECTED] Sent: Friday, November 29, 2002 7:26 PM Subject: [JAVA3D] Efficiency (follow up) My I follow up my own message. The problem was that the same command that updates the location of a surface often takes an unreasonably long time. Here's the code snipet:    for (int i = 0; i < height; i++) {       for (int j = 0; j < width; j++) {         myTA.setCoordinate(i*width + j, surface.get(i, j));         myTA.setNormal(    i*width + j, surf.normal(i, j));       }     } Suppose this is called "Synchronizing". Then below is how much time it takes: Synchronizing: 0.5 Synchronizing: 0.1 Synchronizing: 5.6 Synchronizing: 0.1 Synchronizing: 9.3 Synchronizing: 0.1 Synchronizing: 0.1 Synchronizing: 0.1 Synchronizing: 0.1 Synchronizing: 16.2 Synchronizing: 0.1 Synchronizing: 0.1 Synchronizing: 26.4 Synchronizing: 9.0 Synchronizing: 0.1 Synchronizing: 0.1 Synchronizing: 9.1 Synchronizing: 0.1 Synchronizing: 8.9 Synchronizing: 0.1 Synchronizing: 8.4 Synchronizing: 9.1 Synchronizing: 0.1 Synchronizing: 0.1 Synchronizing: 4.1 Synchronizing: 0.1 Synchronizing: 13.3 Synchronizing: 0.1 Synchronizing: 0.1   My such huge discrepancy? How do I avoid it. I'll take a kluge at this point...   Thanks   Bura Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now