game engine why not like compositor?

we blenderers have a beautiful node based compositor for our images but why we do not have this for our game engine.

if i work with the system now i can’t make any complicating codes because after 10 of them its like a spiderweb.

if we would use nodes for our game engine like our compositor you could make it more organized and code better.

also i see a lot of other game engines also use such a system.
i hope some kind of developer is seeing the potential in it.
we need a better system.

I’m not exactly sure what you’re referring to - the Node editor does work in the BGE, while the Compositor doesn’t. I think you can change properties of as well as distort images through the Node editor, however.

The compositor doesn’t work in the game engine because it generally takes about a second or more to render its effects- that would mean you’d get a less than 1 fps. It’s a very flexible system for post processing, but it’s not very efficient. You might be able to write a script that exports node setups into full-screen GLSL filters, though I don’t really know it that’s possible or how you might go about doing it.

The compositor wouldn’t make too much sense in the game engine anyway from what I can tell. Even games that do have effects, they’re almost always global effects (DOF, motion blur, bloom, etc). Since elements in a game are constantly changing, I don’t see much point in using the compositor. Adding effects to the post processes would make sense to me however. What type of compositing are you trying to do in the game engine?

~David

I think he is referring to nodal logic…

Oh. Well, that’s an idea that was proposed for implementation, but I don’t think it ever really took off.

Yeah I believe he’s talking about logic bricks… which I do agree would be alot easier to keep organized if it were set up like the compositor

yes i mean the logic bricks
i saw thaty the unreal engine works like our compositor.
with nodes.

i do not mean the graphical effects but the programming.

if i work with bge now it quickly looks like a spiderweb.
the system for our compositor gives you more feedom and lets you place the nodes as you want.
if we do this also for our bge we can make longer codes and keep track of our brains.
http://studentpages.scad.edu/~jmilla20/KisMet.jpg

I understand & I agree.

“Fixed” location logic bricks are sometimes looks messy when 1 sensors connected to multiple controllers & 1 controller connected to multiple actuators… especially when other object’s controller connected to other object’s actuators.

You can always try to code your own filters using the BGE’s GLSL functions, then you make a large script where everything is controlled by properties and all you need then is a second script to list the conditions and requirements for them to change (because most people do not put a lot of extra game logic inside 2D filter scripts).

There’s a lot of 2D filter scripts by Martinsh to help you get started (though his are less in the way of just doing basic things like sepia filtering)

It’s not a nodal approach by any means, but it can avoid a big mess of bricks.

Hierarchical Nodal Logic for Blender 2.5
http://wiki.blender.org/index.php/Dev:Source/GameEngine/NodalLogic