-----Urspr�ngliche Nachricht-----
Von: Rob Nugent [mailto:[EMAIL PROTECTED]
Gesendet: Donnerstag, 19. Juni 2003 15:32
An: [EMAIL PROTECTED]
Betreff: Re: [JAVA3D] AW: [JAVA3D] AW: [JAVA3D] Stress testing and
'buildGA' cost.


>I think that my point was that the *API* of Java3D is currently independent
of
>the use of OpenGL or Direct3D. Your suggestion of changing Canvas3D to
extend
>GLCanvas wouldn't hang together for Direct3D unless the API got 'split',
which
>seems like a very bad idea.

Well yes, you're right, but that was only a first (and overenthusiastic)
suggestion. You can always deal with that. Let's assume that we can use the
delegation model. Canvas3D may be a lightweight awt/swing component
containing inside a heavyweight GLCanvas or whatever 3D canvas might be
used. *Then* it might have a get3DCanvas method which may return a Canvas
object, which can be an instance of GLCanvas or D3DCanvas or whatever. That
way you can shield the user from the underlying API, and at the same time
giving him the power to access it if he wants. And you don't have to split
the API. And I'm pretty sure that there are many more other ways(patterns)
to allow the user the access to low level specific api while keeping the
independence of the high level api.

Cheers,

Florin

==========================================================================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