Irrlicht Engine

Im always though that it isnt performance wise to split model to multible meshes.
I dont think irrlicht frustum cull every mesh, but models, but still.
Correct if Im wrong?

If using bounding box, ie car, which is only one model. There is lower-part and upper-part which should use different collision shapes (because upperpart isnt as wide as lower part (when looking at side view)). So there one must still create these manually (or using per-poly collision, dont remember...

I havent done this by myself but I might create collision shapes with blender/3dsmax and name these COLL_* and then parse them in code, and create shapes from them. I think with boxes this works, and maybe spheres but ellpisoids I dont know. Or get these dimensioins in blender and put them in collis...

Some code from my unfinished def renderer (so you get the idea). It creates 3 RT and renders all of them at once. init:   irr::core::array<irr::video::IRenderTarget> multiRenderTarget;       colorRT = driver->addRenderTargetTexture(core::dimension2d<u32>(w,h), "colorMap", ECF_A8R8G8B8);   ...

I didn't see a way of drawing the scene and at the same time drawing it's depth map to another texture.

You can do that using MRT.

  Can i still rotate, move and scale the individual nodes even if their meshes is combined or how does this work?   No. When using CMeshCombiner, it creates 1 object which contains all data.   Can i remove one of them later on or do i need to redo the combination then?   Cant remove invidual object...

But i guess it helps in drawing models as well... (As seen with the dwarfs). You can batch static models, you cant animate batched models. Got fps 150 -> 300 when batched my grass.           array<IMeshSceneNode*> nodes;         for(int q=0;q<10000;q++)         {             const float S=50;      ...

Noticed that               if(collMananger->getCollisionPoint(ray, selector, point, triangle, node))             {                 sphere.getNode()->setPosition(point);                 printf("> %s \n", node->getName());             }   gives sometime right and sometime wrong node-name. Sp...

Oops somehow I missed that one, now it works right. Thanks.

If I create, say bush, with planes (with image of leaves, includes alpha channel), I want to read+write to zbuffer but only where alpha!=0. Havent succeeded with this yet. Current code:           sceneManager->getParameters()->setAttribute(ALLOW_ZWRITE_ON_TRANSPARENT, true);           ITexture *imag...

Can I set shader parameters without using callback class?

Thanks to both.
After unsuccesful patching, I made MyNodeManager from ILightManager and now know how to continue my def renderer.

Can I filter objects somehow, what I want to render at pass 1, and then render others at pass 2? Ie, with deferred renderer, I dont want render transparent objects to gbuffer, I want render them later. And "light spheres" must be invisible too, when rendering gbuffer. Or how this should be...

Here is something:
http://irrlicht.sourceforge.net/forum/v ... 0&start=15

But if you dont like to patch your codes / if your scene is static, you can use MeshCombiner
http://irrlicht.sourceforge.net/forum/v ... =6&t=39598

I compiled irrlicht sources from trunk, with Cg enabled. And installed cg toolkit 3.1 sdk. Now when I use D3D9 and exit, comes exception. With OGL, exiting works. Test code which does just that:   int main() {     E_DRIVER_TYPE driverType = EDT_DIRECT3D9; //   Throws exception when exiting     //E_D...