Easily confused

> I am having a terrible time trying to get my physics demo working.  I
> have a TAvatarReplica subclass that is the primary physics particle
> type for now.  I have destroyed its inherited downPointer and added a
> forwardPointer, which I intend to have pointing in the direction of
> the particle movement.  I have tried a couple of different approaches
> to making this work over the past few weeks and so far been
> unsuccessful.
>
> I have also subclassed TRay and made it visible only for the purpose
> of making the default TFrame #renderAxes: z-pointing arrow bigger than
> the other two, hopefully indicating the direction of the TRay picking.
>  XYZ-RGB, I like it!
>
> In the class comments for TRay, it seems that the best approach should
> be to make the TRay a child of the avatarReplica with a zero
> translation, but I am having difficulty figuring out how to separate
> the child direction-pointing from the parent facing-direction, which
> is necessary because the direction of travel and the facing direction
> are really separate things (in this world anyway).  Is there maybe a
> nice example of this somewhere?
>
> Oh, did I mention that I also replaced all the mouse and keyboard
> interaction so that #driveStep always causes the mouse to only change
> the viewing direction and got rid of #fall altogether as the
> replica/particle movement is always calculated in a future stepping
> loop within the TSpace subclass (in #primitiveStep).  I'm using the
> WASD keys and arrow keys to add acceleration.
>
> Anyway, the other alternative that I apparently wasted lots of time on
> was to make the TRay a child of the TSpace instead of the
> avatarReplica and just make its translation equal the translation of
> the avatar/particle.  This simplified the calculation of the facing
> direction, or at least I thought it should.  Sometimes the TRay seemed
> to point in the direction I expected, sometimes it seemed to point in
> the opposite direction, and sometimes some other unrelated direction.
> Weird!  And this was supposed to be the simple case!  This prompts the
> question "in which direction does a TRay test?"  Does it pick things
> in a negative Z direction relative to where it's facing?  This is what
> I would expect.  Or in a relative positive Z direction?  And how does
> this relate to the default TFrame #renderAxes: ?  Does the blue arrow
> point in the direction that the TRay is picking in?
>
> Any guidance would be appreciated.
>
> Trying to get unconfused,
> Dave
>