On Fri, 6 Jul 2001, Yazel, David J. wrote:
> 2. I don't understand why it would be that much faster. The GeometryUpdater
> is just a basic sychonization protection.
Just shooting blindly (I also have no problems with speed of
GeometryUpdater):
Maybe rendering pipeline has to be stopped and wait for updater to
finish ? It would be obvious to have geometry updaing running in other
thread, so data would be ready for renderer. But maybe internal order
of stuff in this case is very bad and renderer has to stop at every shape
(this would explain degration of performance with more shapes). If you
will add big cost of thread switch and possibly some strange graphic
card/driver...
As for the updating without GeometryUpdater... results are undefined. It
is perfecly ok for some implementation of java3d to not display changes at
all, if some data has to be copied out of by-ref array.
Artur
===========================================================================
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".
