Hi Justin..
Thanks for your reply...
I thought that immediate mode might be the answer.. I'm getting okay
performance with renderOnce(), so I'll probably stick with it for now.
I've ended up writing a method to remove the Link, SharedGroup and excess
Groups manually... It's given a substantial speedup... The number of nodes
in my tree reduces from approximately 200 to 60 without any change in the
model.. Not a general solution but it works fine for me :)
As far as Hanim goes.. I can't really wait.. I've had to add my own loader
to read all the joint information. My real problem is animating it. I've
ended up implementing a basic key frame animator. Behaviours won't mix well
with the rest of my code. There must be someone out there doing H-Anim
animation using Java3D? If not maybe someone will be interested in what I'm
doing, or does everyone use EAI and Vrml browsers?
Cheers
John
----- Original Message -----
From: "Justin Couch" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Thursday, June 07, 2001 1:17 AM
Subject: Re: [JAVA3D] Performance
> John Sear wrote:
>
> > 1.) I want to explicitly state when to render new frames. I'm
currently
> > using stopView() and renderOnce()... but the javadoc states these are
> > heavyweight and should only be used for verification... What's the
preferred
> > method?
>
> This sortof sounds a bit like you need to use immediate mode rendering.
> I occasionally use renderOnce - particularly in image capture.
> Sometimes, if there is no movement or animation in the scene, the canvas
> will not render. Considering that you need a call in postSwap() to read
> the buffers, you need to force another frame render.
>
> > 2.) I'm using VRML97 loader and end up with a large number of nodes
which
> > basically do nothing... i.e. Groups with a single child, link and
> > sharedgroups when the data isn't shared. (I think this is due to DEF in
the
> > VRML source). Anyway, should I be removing these manually or does
compile
> > automatically spot this?
>
> Yes, this is a really hard problem in mapping VRML to Java3D. We've
> re-implemented the code twice now and will probably have to do it again.
> The reason for the shared groups is that it is possible that someone
> might trawl the scenegraph with the EAI or a script and dynamically add
> items to the scenegraph using direct node references. This makes things
> *really* sticky to deal with and so we have to build all these other
> ancillary structures. It is annoying to say the least.
>
> > P.S. If anyone has any good information for using h-anim with Java3D.
I've
> > found nothing :( and am designing everything myself.
>
> AFAIK, there is nothing. The Xj3D loader is currently working towards
> that goal for Siggraph. First goal is Java in the scripts, and then we
> need to support Protos. Proto support is about 50% done, but is on the
> back-burner as we think scripts are more important. If you can wait for
> a couple of months, then we should have decent, real proto and script
> node support.
>
> --
> Justin Couch http://www.vlc.com.au/~justin/
> Freelance Java Consultant http://www.yumetech.com/
> Author, Java 3D FAQ Maintainer http://www.j3d.org/
> -------------------------------------------------------------------
> "Humanism is dead. Animals think, feel; so do machines now.
> Neither man nor woman is the measure of all things. Every organism
> processes data according to its domain, its environment; you, with
> all your brains, would be useless in a mouse's universe..."
> - Greg Bear, Slant
> -------------------------------------------------------------------
>
>
===========================================================================
> 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".