Hi Kelvin,
That's it :)
The bug-description is EXACTLY what is my problem.
Thx great for your patience and your help (i see, my descriptions were a bit
confusing ;).
The sad part about all that is, as i feared, there is no workaround (the
customer workarround is'nt satisfactory, its even impracticable for my
application) and not even a evaluation. This is too bad, because its a IMHO crucial
point in interactive environments to allow objects be translated more then some
thousands kilometers without causing jitter
(considering the sufficient precision of double and even float).
I would appreciate if you could have a look at this bug and try to point me
to a solution or workaround, i have no more ideas how to tackle this.
And another question is, when (or under what conditions) bugs are evaluated?
Thx in advance
- emanuel
> Hi Emanuel,
>
> The previous reply is based on the code in the mail. I didn't
> go back the archive and search the original email.
>
> Here is from the original email:
>
> if(running&&(theCriteria instanceof WakeupOnElapsedTime))
> {
> viewTransform.mul(moveTrans);
>
> viewTransform.get(viewTranslation);
>
> StarsTransform.setTranslation(viewTranslation);
>
> StarsTransformGroup.setTransform(StarsTransform);
> viewTransformGroup.setTransform(viewTransform);
> }
>
> This does not apply to both cases. Since
> the workaround mention in bug is
>
> public void processStimulus(Enumeration criteria) {
> Modify Object transform as in RotationInterpolator
> Modify Camera View point using object transform
> }
>
> But the above code snippet does not set view transform
> base on object transform so I guess your code is ok.
>
> In the original email you mention:
>
> "The strange thing about this is that this problem only shows up, when
> the translational part of the transforms becomes larger."
>
> Check out if the following bug match:
> 4760871 - Jitter when objects are translated far from origin.
>
>
> - Kelvin
> ---------------
> Java 3D Team
> Sun Microsystems Inc.
>
>
> Emanuel Brzostowski wrote:
> > Hi Kelvin
> >
> > Slowly i start to get mad ;)
> > here is why:
> > You wrote:
> >
> >>Besides, I'm not sure why you need to get
> >>SchedulingInterval back and do different things,
> >>if you are really doing this in single behavior it should be
> >>
> >> viewTransform.get(viewTranslation);
> >> StarsTransform.setTranslation(viewTranslation);
> >> StarsTransformGroup.setTransform(StarsTransform);
> >> viewTransformGroup.setTransform(viewTransform);
> >>
> >>instead of
> >>
> >> if(getSchedulingInterval()==1)
> >> {
> >> viewTransform.get(viewTranslation);
> >> StarsTransform.setTranslation(viewTranslation);
> >> StarsTransformGroup.setTransform(StarsTransform);
> >> }
> >> else
> >> {
> >> viewTransformGroup.setTransform(viewTransform);
> >> }
> >
> >
> > The code you suggested first, was exactly that, what i tried first - to
> no
> > avail.
> > I also modified the stimulus for the behavior to be
> > WakeupOnEllapsedFrames(0) - No sucess.
> >
> > To clarify my problem, i first include a quote, taken from the ApiDoc
> about
> > the
> > Behavior class:
> >
> > "[...]A typical behavior will modify one or more nodes or node
> components in
> > the scene graph. These modifications can happen in parallel with
> rendering.
> > In general, applications cannot count on behavior execution being
> > synchronized with rendering. There are two exceptions to this general
> rule:
> >
> > 1.) All modifications to scene graph objects (not including geometry
> > by-reference or texture by-reference) made from the processStimulus
> method
> > of a
> > single behavior instance are guaranteed to take effect in the same
> rendering
> > frame.
> >
> > 2.) All modifications to scene graph objects (not including geometry
> > by-reference or texture by-reference) made from the processStimulus
> methods
> > of the
> > set of behaviors that wake up in response to a WakeupOnElapsedFrames(0)
> > wakeup condition are guaranteed to take effect in the same rendering
> > frame.[...]"
> >
> > The "workaround" suggested for the (non-)bug #4332793 clearly references
> > Statement 2.)
> >
> > But the problem i have, obviously references Statement 1.) !
> > (Two transforms in ONE behavior, not reflected in the same frame).
> > Plz have a look at the code-snippets in my first posting, you will see
> that
> > the premises
> > of Statement 1.) are met in my case. So the conclusion have not to be be
> > false,
> > otherwise the implication in Statement 1.) is proven to be WRONG.
> > I believe, the testcase in my first posting proves this and thus, there
> > seems to be another bug with the Behavior.
> >
> > It would be nice, if you could check this again, and hopefully proving
> me
> > wrong
> > (and offer a solution if so) ;)
> >
> > - emanuel
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >>Hi Emanuel,
> >>
> >> The workaround and new API apply when several
> >>behaviors wakeup EVERY frame. i.e. several behavior
> >>use wakeupOnElapsedFrame(0) and we need to order them
> >>appropriately.
> >>
> >> However in your case there is two behaviors,
> >>wakeupOnElapsedTime(10) which postId(0) to trigger another
> >>wakeupOnBehaviorPost() so above things will not apply here.
> >>
> >>
> >>Besides, I'm not sure why you need to get
> >>SchedulingInterval back and do different things,
> >>if you are really doing this in single behavior it should be
> >>
> >>
> >>
> >>
> >> viewTransform.get(viewTranslation);
> >> StarsTransform.setTranslation(viewTranslation);
> >> StarsTransformGroup.setTransform(StarsTransform);
> >> viewTransformGroup.setTransform(viewTransform);
> >>
> >>
> >>instead of
> >>
> >> if(getSchedulingInterval()==1)
> >> {
> >> viewTransform.get(viewTranslation);
> >> StarsTransform.setTranslation(viewTranslation);
> >> StarsTransformGroup.setTransform(StarsTransform);
> >> }
> >> else
> >> {
> >> viewTransformGroup.setTransform(viewTransform);
> >> }
> >>
> >>- Kelvin
> >>
> >>
> >>Emanuel Brzostowski wrote:
> >>
> >>>Hi Kelvin
> >>>
> >>>
> >>>
> >>>> Check out evaluate of bug
> >>>>4332793 -
> >>>> View transform & object transform not synchronize in a single
> behavior
> >>>
> >>>
> >>>I already found this bug in database before, but it is not exactly what
> >>
> >>i
> >>
> >>>need.
> >>>In the description of the problem in the bug above two behaviors are
> >>>involved
> >>>where both should be executed in the same time.
> >>>But ok, i tracked it down to find the possible soultion (forcing the
> >>
> >>order
> >>
> >>>of first running the interpolator and after that running the
> >>
> >>camera-behavior
> >>
> >>>by setting the sheduling intervalls).
> >>>So i modified the testcase (in a IMHO strange way), ensuring that my
> >>
> >>object
> >>
> >>>transform
> >>>is executed before the view-transform.
> >>>I've done this with another two behaviors wich react on the same
> stimuli
> >>>(WakeupOnBehaviorPost) and so wake up the same time.
> >>>These behaviors got sheduling-intervalls to reflect their priority
> >>
> >>(first
> >>
> >>>update transform of StarsTransformgroup and secondary update the
> >>>view-transformgroup).
> >>>BUT same problem as before: rendered frame does'nt shows up both
> >>
> >>transforms
> >>
> >>>:(
> >>>
> >>>the following is some of the code i produced (unchanged parts cut out):
> >>>plz have a look, its short ;)
> >>>===============================
> >>>.
> >>>.
> >>>.
> >>> Canvas3D canvas3d = new Canvas3D(null);
> >>> Transform3D viewTransform;
> >>> TransformGroup viewTransformGroup;
> >>> Transform3D StarsTransform;
> >>> TransformGroup StarsTransformGroup;
> >>> View myView;
> >>> timedBehavior behav;
> >>>.
> >>>.
> >>>.
> >>> public BranchGroup buildViewBranch(Canvas3D c)
> >>> {
> >>> .
> >>> .
> >>> .
> >>> // this behavior posts the id
> >>> behav = new timedBehavior();
> >>> behav.setSchedulingBounds(bounds);
> >>> viewTransformGroup.addChild(behav);
> >>>
> >>> // following behaviors wake up on the posted id
> >>> calledBehavior LowSIBehav = new calledBehavior();
> >>> LowSIBehav.setSchedulingBounds(bounds);
> >>> LowSIBehav.setSchedulingInterval(1);
> >>> viewTransformGroup.addChild(LowSIBehav);
> >>>
> >>> calledBehavior HighSIBehav = new calledBehavior();
> >>> HighSIBehav.setSchedulingBounds(bounds);
> >>> HighSIBehav.setSchedulingInterval(5);
> >>> viewTransformGroup.addChild(HighSIBehav);
> >>> .
> >>> .
> >>> .
> >>> return(viewBranch);
> >>> }
> >>>.
> >>>.
> >>>.
> >>> public class timedBehavior extends Behavior
> >>> {
> >>> public boolean running = false;
> >>> WakeupCriterion[] wakeConditions;
> >>> WakeupOr oredConditions;
> >>> AWTEvent ae[];
> >>> Transform3D moveTrans;
> >>> Vector3d viewTranslation;
> >>>
> >>>
> >>> public void initialize()
> >>> {
> >>> moveTrans = new Transform3D();
> >>> moveTrans.set(new Vector3d(0.0, 0.0, 10000.0));
> >>> viewTranslation = new Vector3d();
> >>>
> >>> wakeConditions = new WakeupCriterion[3];
> >>> wakeConditions[0] = new WakeupOnAWTEvent(KeyEvent.KEY_PRESSED);
> >>> wakeConditions[1] = new WakeupOnAWTEvent(KeyEvent.KEY_RELEASED);
> >>> wakeConditions[2] = new WakeupOnElapsedTime(10);
> >>> oredConditions = new WakeupOr(wakeConditions);
> >>> wakeupOn(oredConditions);
> >>> }
> >>>
> >>>
> >>> public void processStimulus(Enumeration criteria)
> >>> {
> >>> WakeupCriterion theCriteria;
> >>> if(criteria.hasMoreElements())theCriteria =
> >>>(WakeupCriterion)criteria.nextElement();
> >>> else
> >>> {
> >>> wakeupOn(oredConditions);
> >>> return;
> >>> }
> >>>
> >>> if(theCriteria instanceof WakeupOnAWTEvent)
> >>> {
> >>> ae = ((WakeupOnAWTEvent)theCriteria).getAWTEvent();
> >>> if(ae[0].getID()==KeyEvent.KEY_PRESSED) running = true;
> >>> else if(ae[0].getID()==KeyEvent.KEY_RELEASED) running = false;
> >>> }
> >>> else
> >>> {
> >>> if(running&&(theCriteria instanceof WakeupOnElapsedTime))
> >>> {
> >>> viewTransform.mul(moveTrans);
> >>> this.postId(0);
> >>> }
> >>> }
> >>> wakeupOn(oredConditions);
> >>> }
> >>> }
> >>>
> >>>
> >>> public class calledBehavior extends Behavior
> >>> {
> >>> WakeupCondition wakeCondition;
> >>> Vector3d viewTranslation;
> >>>
> >>> public void initialize()
> >>> {
> >>> viewTranslation = new Vector3d();
> >>> wakeCondition = new WakeupOnBehaviorPost(behav, 0);
> >>> wakeupOn(wakeCondition);
> >>> }
> >>>
> >>>
> >>> public void processStimulus(Enumeration criteria)
> >>> {
> >>> //ensure, behavior where this condition is true runs first
> >>> if(getSchedulingInterval()==1)
> >>> {
> >>> viewTransform.get(viewTranslation);
> >>> StarsTransform.setTranslation(viewTranslation);
> >>> StarsTransformGroup.setTransform(StarsTransform);
> >>> }
> >>> else
> >>> {
> >>> viewTransformGroup.setTransform(viewTransform);
> >>> }
> >>> wakeupOn(wakeCondition);
> >>> }
> >>> }
> >>>===============================
> >>
> >
> > --
> > +++ GMX - Mail, Messaging & more http://www.gmx.net +++
> > NEU: Mit GMX ins Internet. Rund um die Uhr f�r 1 ct/ Min. surfen!
> >
> >
>
===========================================================================
> > 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".
>
>
===========================================================================
> 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".
>
--
+++ GMX - Mail, Messaging & more http://www.gmx.net +++
NEU: Mit GMX ins Internet. Rund um die Uhr f�r 1 ct/ Min. surfen!
===========================================================================
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".
