This is another bug hunter program combined with something that I am
trying to do that is not quite working right. I have information about the
simulation that I am wanting to display in the simulation that is
constantly changing.
One solution is to use Text2D's a short distance from the camera and to
constantly update those. What I tried to do instead though is subclassed
Canvas3D and made it a pump for swing's state change events. Then I push
out a state change after each postSwap and the listener for that draws on
the canvas the appropriate info. This cuts out the overhead of placing the
Text2D's in the scene and also rerending the info over and over. (In the
actual application of this process the info is changing nearly every
frame.)
The problem is that there is flicker with the text. I am assuming this is
a delay beween the swap and the drawing if the string. Does anyone have
any suggestions on how to fix this?
The actual bug though had to do with long distance translations in the 1.2
implementation. Somewhere around 100000 meters the elements in shapes
being transformed are translated different distances (cone ends and bodies
for instance.) It gets progressively worse as the distance
grows. Eventually when dealing with stuff in the stratosphere like missile
simulations (~6000000 m) objects are hardly recognizable.
I briefly described this program in a comment at the beginning and you can
look at that if you want to know how it works. Oh yeah, it is
jarred (so as to include a texture map.) =) The comment is:
/**
* This is a little test program to show a bug in the implementation of
* the 1.2 api. As the distance increases beyond ~100000 meters the
* translations of different shapes become non-uniform. This does not
* occur in the 1.1 implementation.
*
* As always I could not be contented not to play a bit so the scene
* is that of a satellite orbiting the earth. It is possible to texture
* map the earth using the -t argument on the command line. Also the s
* key will sync the earth and orbit speeds. The v key will dump the
* current status of program variables. Left, right, up, down, page up,
* page down, home and end will change the distance with increasing
* magnitudes of change.
*
* Another thing that I am curious about is the behavior of the rotation
* interpolators when the alpha has an increasing duration of 1 or two.
* It doesn't look right to me though that might just be me. =)
*
* 2000/09/25 -- wjh
*/
Any help on the flicker appreciated.
Will Holcomb
TranslationTest.jar
