I'm trying to get a measure of the impact of using a non-passive WakeupOnElapsedFrames WakeUp criterion on instantiating swing widgets such as JColorChooser and JFileChooser. It appears that regardless of the framecount number (0 - 1000, for example) the creation times of such widgets could be less than a second (passive criterion) to atleast 6 seconds (non-passive criterion) on a WinXP P4 1.6GHz system.
Attached is a simple test case to examine the impact. The testcase has a scenegraph with a WakeupOnElapsedFrames behavior, no geometries, and a simple GUI with a JButton which upon clicking will invoke a JColorChooser. Change the passive parameter flag of the wakeup criterion to examine the impact. The differences in JColorChooser invocation times are as reported above. Is this expected behavior ? The impact appears to be primarily on creation times of such widgets and not on the overall responsiveness of the Swing GUI based on my experiments with a fairly large application. Well, I'm aware that the non-passive criterion keeps the renderer busy, but still a little surprised by its effect. If the above feature is by design, then I may have to modify my strategy of lazily instantiating swing widgets. Comments will be very appreciated. Vaidya =========================================================================== To unsubscribe, send email to [EMAIL PROTECTED] and include in the body of the message "signoff JAVA3D-INTEREST". For general help, send email to [EMAIL PROTECTED] and include in the body of the message "help".
