Hi all,

i just quickly looked through the specification for version 1.3 and
would like to add some comments to it:

1) Input devices and Behaviors:

Extend the API to support also force feedback ?

Define what kind of transformations shall be returned in a sensor:
absolute or difference transformations ? I would prefer difference
transformations. Then implementers of the InputDevice interface and
implementers of behaviors know what kind of input they can expect and
input devices are interchangeable.

Why don't you use the input device interface for mouse input ?
Some design issue: It would be nice to capsulate the function of
MouseRotator, MouseTranslate, OrbitBehavior into classes. Then only one
behavior must be used, but the function depends which class is used in
the behavior. Thus other input deviceses than mice don't have to
implement these functionalities.

By the way: I think implementing the comm api for all platforms which
have a java3d implementation (Linux) would be appreciated.

Extend the Behavior API with a method to obtain the View which has
currently focus (or had focus, if no Canvas3D has currently focus).
Easier to obtain the local coordinate system of the associated
viewplatform to apply transformation in this local coordinate system.

2) Transform3D and Matrix3x

Extend the API with a transformTranspose method which transforms a
vector with the transposed matrix. Good for doint coordinate
transformation into a different coordinate system and back into the
original coordinate system, when only rotations are considered.

Extend the API for rotations to accept also the sin and cos of an angle.
Sometimes they an be calculated quite easy and there is no need to
convert them into angles to use the methods.

3) Stereoscopic images

Some more comments on stereoscopy in the documentation would be helpful.
Right now we are a minority. But how long ?

There are several formats for stereoscopic images: above below,
interlaced, white code, left/right image buffers, ... .
As far as I see through all of this java3d supports stereo through the
underlying API (e.g. open GL left/right buffers, ? for direct3d). Right
?
But there are these other formats which are not supported, or which have
to be implemented by everybody (e.g. i want to use Sterographics
above/below format) who needs them, e.g. using immediate mode rendering.
Why not let java3d support also this formats (e. .g. some parameter can
be used to determine the format) ?
How do I change the paramters for the formats like the "focus point" and
others if I can't get to the drivers or graphics API ?

I know this stereo stuff is also a driver stuff, but any hints in the
documentation are welcome.

Ok, that's it for now.

regards
Paul

PS: Some extensions of the API look very promising. Can't wait to work
with it.

===========================================================================
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".

Reply via email to