David Yazel wrote:

[ a good rant ]

Dave, I think much of your problems could be attributed to looking in
the wrong place. I am not a game developer by any stretch of the
imagination. My experience is in simulation and visualisation. The
things I hear and occasionally see from game developers ring true with
your statemts of being unwilling to share knowledge. Much of this can be
construed as "commercial advantage" where slick graphics can mean the
difference between a game which breaks even and one which fails. For
many game developers, they need to keep ahead of the others so that they
can continue to earn a living doing the thing they love rather than
having to hack out VB reports for some bank (or writing EJB middleware
in my case). Most of the sources you sight for information are game sites.

On the other side of the fence are the visualisation and sim guys. For
them, the competition is much less and they tend to be much more open
about techniques. Simulation guys are even more useful because that are
almost all adept at scenegraph based techniques having been fed a good
diet of SGI Performer over the past decade or so. In addition, many of
the visualisation guys are from an acedemic setting, so there is no real
commercial advantage to be gained and therefore also very open.

Many techniques used in gaming came from the simulation community.
Slightly less rigid, and therefore a bit faster, they work their way
into the games about 2-3 years later. Take for example that paper I
posted a link to. They are architectural folks wanting to do realistic
walkthroughs of cityscapes for urban planners. I'd place good money that
in two years time you will see an FPS that has you running around a
cityscape dodging people trying to shoot the bad guys.

For a start, I seriously recommend that you get to the Siggraph
conferences. You can pick up many interesting ideas and techniques
there. People are also much more willing to chat with you about how to
go about implementing things. Compared to the cost of GDC or JavaOne, it
is money very well spent. Most the techniques I've played with over the
years have come from seeing something at siggraph and then having a go
at hacking up myself.

My last point is to not assume that something that was implemented in
opengl cannot be implemented with j3d. Scenegraphs do have a different
view of the 3D world, but consider that most of the techniques you see
in use in games today were originally built on a scene graph system of
some sort. They are not really that hard to shift from immediate mode to
retained mode. Sometimes you have to change your line of thought, but
not always. Of course, the biggest problem areas in implementing these
algorithms are when you start playing with culling algorithms, where we
have no real access to the j3d internals to override that. To take your
one example - multiple texture images as a single texture - I was using
that technique 4 years ago in a VRML system I developed at the time.
Prior to me using it I know that the sim guys had been using it for at
least 2-3 years before that (that's how they implemented the first
flame/smoke systems).

--
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".

Reply via email to