[x3d-public] A note on standardization.

[Alan Grimes]

Hey, I want to support any previous comments about the caution that must
be applied toward standardization.

First, we must consider the immediate future of VR. Take a look at
Skyrim (or later) and Battlefield 4 as examples of where VR already is
today and get an idea of where it is going in the future.

So therefore standards should do everything possible to avoid stifiling
innovation and should only concentrate on what CAN'T be programmed from
first principles.


So what should be (if it is not already) standardized?

1. Input methods. -- There needs to be a framework that allows hardware
and OS vendors to provide UI's based on
gamepads/keyboards/mice/touchscreens/head trackers, and it needs to
provide application developers a way to map input concepts to
appropriate subroutines. Such as "move forward" and stuff, and maybe
some mechanism for programatically defining and mapping custom input
actions.

( future = neural interface; AI avatar interface; ie a radically
different technology)


2. Hardware accelerated physics; ability to define colliders and note
that the client is expected to perform physics calculations in such a
way, ie specification of gravity and other forces.. Obnoxious example:
Kerbal Space Program and Space Engineers.
Flags would be "Detect" -- run script if collision; "physics" -- take
collision surface into account for physics, and the other parameters
such as mass and such...   I haven't studied the existing standard
enough to figure it out...



I feel that it should be possible to do other things such as
client-server and multiuser through ECMA script, embedded JAVA, and the
usual back-end servers... I'm kinda partial to the Tomcat /j2ee stack...
others use "Node.js" which kinda sticks in my craw because I have a dim
view of ECMA script... So there isn't much need for standardization
because infrastructure for this kind of thing is already commonly
available...


-- 
IQ is a measure of how stupid you feel.

Powers are not rights.